Skip to content

Conversation

@avik-pal
Copy link
Collaborator

@avik-pal avik-pal commented Jan 6, 2026

No description provided.

@avik-pal avik-pal force-pushed the ap/scatter_gather_simplify branch from bdf7779 to c89d3ad Compare January 6, 2026 21:02
Base automatically changed from ap/scatter_simplify_add_sub to main January 6, 2026 23:20
@avik-pal avik-pal force-pushed the ap/scatter_gather_simplify branch from c89d3ad to 00f0ff5 Compare January 7, 2026 00:16
@avik-pal avik-pal marked this pull request as ready for review January 7, 2026 00:17
@avik-pal avik-pal requested a review from wsmoses January 7, 2026 00:17
@avik-pal avik-pal merged commit fe27139 into main Jan 7, 2026
16 of 17 checks passed
@avik-pal avik-pal deleted the ap/scatter_gather_simplify branch January 7, 2026 02:16
Copy link
Contributor

@github-actions github-actions bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

EnzymeJAX Benchmarks

Details
Benchmark suite Current: 00f0ff5 Previous: 99d2b63 Ratio
actmtch / JaXPipe / cpu / Primal 0.000006931539996912761 s 0.000007074660015859991 s 0.98
actmtch / Jax / cpu / Primal 0.000006383279994679469 s 0.00000704749997566978 s 0.91
actmtch / HLOOpt / cpu / Primal 0.000007861699991735804 s 0.000007264960004249588 s 1.08
actmtch / PartOpt / cpu / Primal 0.000006531280007493479 s 0.00000608861999353394 s 1.07
actmtch / IPartOpt / cpu / Primal 0.000006892940014040505 s 0.000006828119985584636 s 1.01
actmtch / DefOpt / cpu / Primal 0.000007686800004194083 s 0.000008983800016721943 s 0.86
actmtch / IDefOpt / cpu / Primal 0.000007646740000382124 s 0.00000724573998923006 s 1.06
actmtch / JaXPipe / cpu / Forward 0.000011129840002013224 s 0.000011053500002162763 s 1.01
actmtch / Jax / cpu / Forward 0.000009963620002508832 s 0.000009771300019565388 s 1.02
actmtch / HLOOpt / cpu / Forward 0.000011202019986740196 s 0.000011151759999847854 s 1.00
actmtch / PartOpt / cpu / Forward 0.000011049059996821598 s 0.00001049553998200281 s 1.05
actmtch / IPartOpt / cpu / Forward 0.000011706860000231244 s 0.00001088953998987563 s 1.08
actmtch / DefOpt / cpu / Forward 0.00001114062000397098 s 0.000010273019979649687 s 1.08
actmtch / IDefOpt / cpu / Forward 0.000010398760009593388 s 0.000010934499978247913 s 0.95
actmtch / JaXPipe / cpu / PreRev 0.000010575580008662656 s 0.000011111580015494835 s 0.95
actmtch / JaXPipe / cpu / PostRev 0.000011078560009991634 s 0.000009981279990824987 s 1.11
actmtch / JaXPipe / cpu / BothRev 0.00001184106000664542 s 0.000011585639986151364 s 1.02
actmtch / Jax / cpu / BothRev 0.000009486260007633972 s 0.000009328440046374454 s 1.02
actmtch / HLOOpt / cpu / PreRev 0.000011369359997388528 s 0.000010915980010395289 s 1.04
actmtch / HLOOpt / cpu / PostRev 0.000013244779995602584 s 0.000012751240028592292 s 1.04
actmtch / HLOOpt / cpu / BothRev 0.000011323799999445329 s 0.000011103559972980291 s 1.02
actmtch / PartOpt / cpu / PreRev 0.00001108467999529239 s 0.00001094761998501781 s 1.01
actmtch / PartOpt / cpu / PostRev 0.000010777779989439296 s 0.000010008060016843955 s 1.08
actmtch / PartOpt / cpu / BothRev 0.000011524839992489431 s 0.000011785139986386638 s 0.98
actmtch / IPartOpt / cpu / PreRev 0.000010764259984625824 s 0.000010656700005711171 s 1.01
actmtch / IPartOpt / cpu / PostRev 0.000010247720001643755 s 0.000010013160017479094 s 1.02
actmtch / IPartOpt / cpu / BothRev 0.00001096026000141137 s 0.000011441900005593195 s 0.96
actmtch / DefOpt / cpu / PreRev 0.000011018639995654666 s 0.0000110185000266938 s 1.00
actmtch / DefOpt / cpu / PostRev 0.000011731099996268312 s 0.00001147651999417576 s 1.02
actmtch / DefOpt / cpu / BothRev 0.000011450160000094912 s 0.000010963859995172243 s 1.04
actmtch / IDefOpt / cpu / PreRev 0.000010706880000270757 s 0.0000109305400201265 s 0.98
actmtch / IDefOpt / cpu / PostRev 0.00001127515999996831 s 0.000011023619990737644 s 1.02
actmtch / IDefOpt / cpu / BothRev 0.000011355779997757054 s 0.000011036960022465792 s 1.03
actmtch / JaXPipe / cuda / Primal 0.0000024 s 0.000002047 s 1.17
actmtch / Jax / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / HLOOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / PartOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / IPartOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / DefOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / IDefOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / JaXPipe / cuda / Forward 0.000010432 s 0.00001008 s 1.03
actmtch / Jax / cuda / Forward 0.000010273 s 0.000010208 s 1.01
actmtch / HLOOpt / cuda / Forward 0.000010464 s 0.000009856 s 1.06
actmtch / PartOpt / cuda / Forward 0.000010112 s 0.000010305 s 0.98
actmtch / IPartOpt / cuda / Forward 0.00001024 s 0.000010176 s 1.01
actmtch / DefOpt / cuda / Forward 0.000010496 s 0.000013472 s 0.78
actmtch / IDefOpt / cuda / Forward 0.000010496 s 0.000010176 s 1.03
actmtch / JaXPipe / cuda / PreRev 0.000010367 s 0.000009856 s 1.05
actmtch / JaXPipe / cuda / PostRev 0.000010656 s 0.000010464 s 1.02
actmtch / JaXPipe / cuda / BothRev 0.000010528 s 0.000010047 s 1.05
actmtch / Jax / cuda / BothRev 0.000010496 s 0.000010016 s 1.05
actmtch / HLOOpt / cuda / PreRev 0.000010688 s 0.000009824 s 1.09
actmtch / HLOOpt / cuda / PostRev 0.0000104 s 0.00001024 s 1.02
actmtch / HLOOpt / cuda / BothRev 0.000010464 s 0.00001008 s 1.04
actmtch / PartOpt / cuda / PreRev 0.000010464 s 0.000010432 s 1.00
actmtch / PartOpt / cuda / PostRev 0.00001072 s 0.000010208 s 1.05
actmtch / PartOpt / cuda / BothRev 0.00001056 s 0.000010272 s 1.03
actmtch / IPartOpt / cuda / PreRev 0.000010496 s 0.000010144 s 1.03
actmtch / IPartOpt / cuda / PostRev 0.000010369 s 0.000010304 s 1.01
actmtch / IPartOpt / cuda / BothRev 0.000010592 s 0.000009792 s 1.08
actmtch / DefOpt / cuda / PreRev 0.000010368 s 0.000010176 s 1.02
actmtch / DefOpt / cuda / PostRev 0.000010528 s 0.000010048 s 1.05
actmtch / DefOpt / cuda / BothRev 0.000010336 s 0.000009728 s 1.06
actmtch / IDefOpt / cuda / PreRev 0.00001056 s 0.000010784 s 0.98
actmtch / IDefOpt / cuda / PostRev 0.000010368 s 0.000010303 s 1.01
actmtch / IDefOpt / cuda / BothRev 0.000010496 s 0.000010144 s 1.03
actmtch / JaXPipe / tpu / Primal 5.63125e-7 s 5.6375e-7 s 1.00
actmtch / Jax / tpu / Primal 5.9635e-7 s 5.9715e-7 s 1.00
actmtch / HLOOpt / tpu / Primal 0.0000021109 s 0.000002100575 s 1.00
actmtch / PartOpt / tpu / Primal 5.95975e-7 s 5.965499999999999e-7 s 1.00
actmtch / IPartOpt / tpu / Primal 5.519500000000001e-7 s 5.52725e-7 s 1.00
actmtch / DefOpt / tpu / Primal 0.0000021587 s 0.0000021609 s 1.00
actmtch / IDefOpt / tpu / Primal 0.00000211555 s 0.000002110575 s 1.00
actmtch / JaXPipe / tpu / Forward 0.00000383495 s 0.000003824075 s 1.00
actmtch / Jax / tpu / Forward 0.000001208875 s 0.00000121035 s 1.00
actmtch / HLOOpt / tpu / Forward 0.0000039393 s 0.000003934275 s 1.00
actmtch / PartOpt / tpu / Forward 0.000003919475 s 0.000003911724999999999 s 1.00
actmtch / IPartOpt / tpu / Forward 0.00000394425 s 0.000003933925 s 1.00
actmtch / DefOpt / tpu / Forward 0.000003917675 s 0.000003913074999999999 s 1.00
actmtch / IDefOpt / tpu / Forward 0.000003933675 s 0.0000039450750000000005 s 1.00
actmtch / JaXPipe / tpu / PreRev 0.00000348315 s 0.0000034795 s 1.00
actmtch / JaXPipe / tpu / PostRev 0.0000016354 s 0.000001645175 s 0.99
actmtch / JaXPipe / tpu / BothRev 0.000003472875 s 0.000003478825 s 1.00
actmtch / Jax / tpu / BothRev 0.0000016383 s 0.000001631225 s 1.00
actmtch / HLOOpt / tpu / PreRev 0.000003474025 s 0.0000034729500000000004 s 1.00
actmtch / HLOOpt / tpu / PostRev 0.00000341205 s 0.000003397575 s 1.00
actmtch / HLOOpt / tpu / BothRev 0.0000034928 s 0.000003481925 s 1.00
actmtch / PartOpt / tpu / PreRev 0.000003413 s 0.0000034036 s 1.00
actmtch / PartOpt / tpu / PostRev 0.000001585075 s 0.000001587925 s 1.00
actmtch / PartOpt / tpu / BothRev 0.00000341385 s 0.000003399525 s 1.00
actmtch / IPartOpt / tpu / PreRev 0.000003467675 s 0.000003505025 s 0.99
actmtch / IPartOpt / tpu / PostRev 0.000001636875 s 0.000001635125 s 1.00
actmtch / IPartOpt / tpu / BothRev 0.0000034779 s 0.00000346705 s 1.00
actmtch / DefOpt / tpu / PreRev 0.0000034049750000000004 s 0.0000034026499999999995 s 1.00
actmtch / DefOpt / tpu / PostRev 0.00000341465 s 0.0000034195 s 1.00
actmtch / DefOpt / tpu / BothRev 0.000003415325 s 0.000003396175 s 1.01
actmtch / IDefOpt / tpu / PreRev 0.000003471525 s 0.00000349265 s 0.99
actmtch / IDefOpt / tpu / PostRev 0.000003397525 s 0.00000341735 s 0.99
actmtch / IDefOpt / tpu / BothRev 0.0000034648250000000003 s 0.000003478575 s 1.00
actmtch / JaXPipe / cpu / Primal 0.000013308 s 0.000007074660015859991 s 1.88
actmtch / Jax / cpu / Primal 0.000013288 s 0.00000704749997566978 s 1.89
actmtch / HLOOpt / cpu / Primal 0.00001399 s 0.000007264960004249588 s 1.93
actmtch / PartOpt / cpu / Primal 0.000013217 s 0.00000608861999353394 s 2.17
actmtch / IPartOpt / cpu / Primal 0.000013468 s 0.000006828119985584636 s 1.97
actmtch / DefOpt / cpu / Primal 0.000013948 s 0.000008983800016721943 s 1.55
actmtch / IDefOpt / cpu / Primal 0.000013969 s 0.00000724573998923006 s 1.93
actmtch / JaXPipe / cpu / Forward 0.00001889 s 0.000011053500002162763 s 1.71
actmtch / Jax / cpu / Forward 0.00001788 s 0.000009771300019565388 s 1.83
actmtch / HLOOpt / cpu / Forward 0.000019075000000000003 s 0.000011151759999847854 s 1.71
actmtch / PartOpt / cpu / Forward 0.000018979 s 0.00001049553998200281 s 1.81
actmtch / IPartOpt / cpu / Forward 0.000018822 s 0.00001088953998987563 s 1.73
actmtch / DefOpt / cpu / Forward 0.000018778 s 0.000010273019979649687 s 1.83
actmtch / IDefOpt / cpu / Forward 0.000019077 s 0.000010934499978247913 s 1.74
actmtch / JaXPipe / cpu / PreRev 0.000019429 s 0.000011111580015494835 s 1.75
actmtch / JaXPipe / cpu / PostRev 0.000017887 s 0.000009981279990824987 s 1.79
actmtch / JaXPipe / cpu / BothRev 0.000019041 s 0.000011585639986151364 s 1.64
actmtch / Jax / cpu / BothRev 0.00001772 s 0.000009328440046374454 s 1.90
actmtch / HLOOpt / cpu / PreRev 0.000018825 s 0.000010915980010395289 s 1.72
actmtch / HLOOpt / cpu / PostRev 0.000019345 s 0.000012751240028592292 s 1.52
actmtch / HLOOpt / cpu / BothRev 0.000019037 s 0.000011103559972980291 s 1.71
actmtch / PartOpt / cpu / PreRev 0.000019119 s 0.00001094761998501781 s 1.75
actmtch / PartOpt / cpu / PostRev 0.0000176 s 0.000010008060016843955 s 1.76
actmtch / PartOpt / cpu / BothRev 0.0000192 s 0.000011785139986386638 s 1.63
actmtch / IPartOpt / cpu / PreRev 0.000018881 s 0.000010656700005711171 s 1.77
actmtch / IPartOpt / cpu / PostRev 0.000017865 s 0.000010013160017479094 s 1.78
actmtch / IPartOpt / cpu / BothRev 0.000019301 s 0.000011441900005593195 s 1.69
actmtch / DefOpt / cpu / PreRev 0.000018928 s 0.0000110185000266938 s 1.72
actmtch / DefOpt / cpu / PostRev 0.000019441 s 0.00001147651999417576 s 1.69
actmtch / DefOpt / cpu / BothRev 0.00001896 s 0.000010963859995172243 s 1.73
actmtch / IDefOpt / cpu / PreRev 0.00001934 s 0.0000109305400201265 s 1.77
actmtch / IDefOpt / cpu / PostRev 0.000019192 s 0.000011023619990737644 s 1.74
actmtch / IDefOpt / cpu / BothRev 0.000019174 s 0.000011036960022465792 s 1.74
add_one / JaXPipe / cpu / Primal 0.000007136839999475342 s 0.000006555000009029755 s 1.09
add_one / Jax / cpu / Primal 0.000006488740009444882 s 0.000008358779950867756 s 0.78
add_one / HLOOpt / cpu / Primal 0.000006747100005668472 s 0.000007089640039339429 s 0.95
add_one / PartOpt / cpu / Primal 0.000006458860004840972 s 0.00000640845996713324 s 1.01
add_one / IPartOpt / cpu / Primal 0.000007376059995749529 s 0.000006578960001206724 s 1.12
add_one / DefOpt / cpu / Primal 0.000007092520004334801 s 0.000006600820006497088 s 1.07
add_one / IDefOpt / cpu / Primal 0.000006892059998335753 s 0.000006797320029363618 s 1.01
add_one / JaXPipe / cpu / Forward 0.000010661379990324347 s 0.000009917019997374155 s 1.08
add_one / Jax / cpu / Forward 0.000010201319998941471 s 0.000009898720018099992 s 1.03
add_one / HLOOpt / cpu / Forward 0.000010804759990605816 s 0.0000102036600401334 s 1.06
add_one / PartOpt / cpu / Forward 0.000010460499997861915 s 0.000010513300003367476 s 0.99
add_one / IPartOpt / cpu / Forward 0.00001002162001668694 s 0.000010385160030637054 s 0.96
add_one / DefOpt / cpu / Forward 0.00001048303999368727 s 0.000010184639986619004 s 1.03
add_one / IDefOpt / cpu / Forward 0.000010665299998890988 s 0.000010276100047121872 s 1.04
add_one / JaXPipe / cpu / PreRev 0.000012086999993243807 s 0.000011770159999286988 s 1.03
add_one / JaXPipe / cpu / PostRev 0.000011386179999135493 s 0.000011907140014955076 s 0.96
add_one / JaXPipe / cpu / BothRev 0.000011940239992327409 s 0.000012559820015667356 s 0.95
add_one / Jax / cpu / BothRev 0.000012094280000383153 s 0.00001256429996828956 s 0.96
add_one / HLOOpt / cpu / PreRev 0.000012364119995709187 s 0.000012178099977973031 s 1.02
add_one / HLOOpt / cpu / PostRev 0.000017246760005491524 s 0.00001428686002327595 s 1.21
add_one / HLOOpt / cpu / BothRev 0.000011854019996917486 s 0.00001226867996592773 s 0.97
add_one / PartOpt / cpu / PreRev 0.000010976239993851775 s 0.00001211154001794057 s 0.91
add_one / PartOpt / cpu / PostRev 0.000011331619991779008 s 0.000012167160066383077 s 0.93
add_one / PartOpt / cpu / BothRev 0.000012177240002984036 s 0.000012089080009900498 s 1.01
add_one / IPartOpt / cpu / PreRev 0.000011616479998792783 s 0.000012506240018410608 s 0.93
add_one / IPartOpt / cpu / PostRev 0.00001173075999986395 s 0.000011844299979202334 s 0.99
add_one / IPartOpt / cpu / BothRev 0.000011182360003658686 s 0.00001200290000269888 s 0.93
add_one / DefOpt / cpu / PreRev 0.000011748839997380857 s 0.000012106419962947256 s 0.97
add_one / DefOpt / cpu / PostRev 0.00001162768000313008 s 0.000011705979941325495 s 0.99
add_one / DefOpt / cpu / BothRev 0.000011909859999832406 s 0.00001240259999576665 s 0.96
add_one / IDefOpt / cpu / PreRev 0.000011416700006066096 s 0.00001211826000144356 s 0.94
add_one / IDefOpt / cpu / PostRev 0.000011974980000104553 s 0.000012116420002712402 s 0.99
add_one / IDefOpt / cpu / BothRev 0.000011627779992977591 s 0.000011627660005615326 s 1.00
add_one / JaXPipe / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / Jax / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / HLOOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / PartOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / IPartOpt / cuda / Primal 0.000002304 s 0.0000019200000000000003 s 1.20
add_one / DefOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / IDefOpt / cuda / Primal 0.000002336 s 0.0000019200000000000003 s 1.22
add_one / JaXPipe / cuda / Forward 0.000010496 s 0.00001024 s 1.03
add_one / Jax / cuda / Forward 0.0000104 s 0.000010368 s 1.00
add_one / HLOOpt / cuda / Forward 0.000010591 s 0.000010208 s 1.04
add_one / PartOpt / cuda / Forward 0.000010208 s 0.00001024 s 1.00
add_one / IPartOpt / cuda / Forward 0.000010464 s 0.000010464 s 1
add_one / DefOpt / cuda / Forward 0.000010271 s 0.000010208 s 1.01
add_one / IDefOpt / cuda / Forward 0.000010336 s 0.000010304 s 1.00
add_one / JaXPipe / cuda / PreRev 0.000025312 s 0.000025728 s 0.98
add_one / JaXPipe / cuda / PostRev 0.000025504 s 0.000025536 s 1.00
add_one / JaXPipe / cuda / BothRev 0.000025344 s 0.000024512 s 1.03
add_one / Jax / cuda / BothRev 0.000025664 s 0.000025568 s 1.00
add_one / HLOOpt / cuda / PreRev 0.000026368 s 0.000025728 s 1.02
add_one / HLOOpt / cuda / PostRev 0.000025632 s 0.00002528 s 1.01
add_one / HLOOpt / cuda / BothRev 0.000025055 s 0.000024737 s 1.01
add_one / PartOpt / cuda / PreRev 0.0000256 s 0.000025152 s 1.02
add_one / PartOpt / cuda / PostRev 0.000025056 s 0.000029312 s 0.85
add_one / PartOpt / cuda / BothRev 0.00002512 s 0.000029568 s 0.85
add_one / IPartOpt / cuda / PreRev 0.000025984 s 0.000029248 s 0.89
add_one / IPartOpt / cuda / PostRev 0.000025664 s 0.000024609 s 1.04
add_one / IPartOpt / cuda / BothRev 0.000025535 s 0.00002448 s 1.04
add_one / DefOpt / cuda / PreRev 0.000026176 s 0.000025377 s 1.03
add_one / DefOpt / cuda / PostRev 0.000025152 s 0.000025152 s 1
add_one / DefOpt / cuda / BothRev 0.000026176 s 0.000024736 s 1.06
add_one / IDefOpt / cuda / PreRev 0.000025951 s 0.000026048 s 1.00
add_one / IDefOpt / cuda / PostRev 0.000026176 s 0.000025536 s 1.03
add_one / IDefOpt / cuda / BothRev 0.00002624 s 0.000025856 s 1.01
add_one / JaXPipe / tpu / Primal 0.0000014275 s 0.0000014431 s 0.99
add_one / Jax / tpu / Primal 0.0000014103 s 0.00000141315 s 1.00
add_one / HLOOpt / tpu / Primal 0.0000014303750000000005 s 0.000001427425 s 1.00
add_one / PartOpt / tpu / Primal 0.000001406225 s 0.0000014064750000000002 s 1.00
add_one / IPartOpt / tpu / Primal 0.000001431325 s 0.000001427325 s 1.00
add_one / DefOpt / tpu / Primal 0.000001405775 s 0.000001403125 s 1.00
add_one / IDefOpt / tpu / Primal 0.000001425125 s 0.0000014284 s 1.00
add_one / JaXPipe / tpu / Forward 0.000001854975 s 0.0000018557 s 1.00
add_one / Jax / tpu / Forward 0.00000185715 s 0.000001852975 s 1.00
add_one / HLOOpt / tpu / Forward 0.000001853375 s 0.000001859325 s 1.00
add_one / PartOpt / tpu / Forward 0.000001853025 s 0.000001843425 s 1.01
add_one / IPartOpt / tpu / Forward 0.000001861825 s 0.000001849025 s 1.01
add_one / DefOpt / tpu / Forward 0.00000184335 s 0.000001850225 s 1.00
add_one / IDefOpt / tpu / Forward 0.000001848375 s 0.000001850875 s 1.00
add_one / JaXPipe / tpu / PreRev 0.0000022364 s 0.000002249275 s 0.99
add_one / JaXPipe / tpu / PostRev 0.0000022371 s 0.000002244275 s 1.00
add_one / JaXPipe / tpu / BothRev 0.000002255675 s 0.0000022332 s 1.01
add_one / Jax / tpu / BothRev 0.0000022505 s 0.0000022404 s 1.00
add_one / HLOOpt / tpu / PreRev 0.00000223845 s 0.000002238025 s 1.00
add_one / HLOOpt / tpu / PostRev 0.000002240025 s 0.00000224105 s 1.00
add_one / HLOOpt / tpu / BothRev 0.000002244075 s 0.0000022355 s 1.00
add_one / PartOpt / tpu / PreRev 0.00000224345 s 0.000002233825 s 1.00
add_one / PartOpt / tpu / PostRev 0.000002243575 s 0.000002244075 s 1.00
add_one / PartOpt / tpu / BothRev 0.000002241375 s 0.0000022386 s 1.00
add_one / IPartOpt / tpu / PreRev 0.0000022426 s 0.0000022358 s 1.00
add_one / IPartOpt / tpu / PostRev 0.000002243225 s 0.0000022391 s 1.00
add_one / IPartOpt / tpu / BothRev 0.00000223725 s 0.000002234875 s 1.00
add_one / DefOpt / tpu / PreRev 0.00000223825 s 0.0000022432 s 1.00
add_one / DefOpt / tpu / PostRev 0.0000022429 s 0.0000022361 s 1.00
add_one / DefOpt / tpu / BothRev 0.0000022437250000000003 s 0.00000224885 s 1.00
add_one / IDefOpt / tpu / PreRev 0.000002232725 s 0.000002238725 s 1.00
add_one / IDefOpt / tpu / PostRev 0.0000022353 s 0.000002251525 s 0.99
add_one / IDefOpt / tpu / BothRev 0.0000022344500000000003 s 0.000002251525 s 0.99
add_one / JaXPipe / cpu / Primal 0.000013259999999999998 s 0.000006555000009029755 s 2.02
add_one / Jax / cpu / Primal 0.000012978 s 0.000008358779950867756 s 1.55
add_one / HLOOpt / cpu / Primal 0.000012773 s 0.000007089640039339429 s 1.80
add_one / PartOpt / cpu / Primal 0.000013019 s 0.00000640845996713324 s 2.03
add_one / IPartOpt / cpu / Primal 0.000012808 s 0.000006578960001206724 s 1.95
add_one / DefOpt / cpu / Primal 0.000012621 s 0.000006600820006497088 s 1.91
add_one / IDefOpt / cpu / Primal 0.000012761 s 0.000006797320029363618 s 1.88
add_one / JaXPipe / cpu / Forward 0.000017661 s 0.000009917019997374155 s 1.78
add_one / Jax / cpu / Forward 0.000017697 s 0.000009898720018099992 s 1.79
add_one / HLOOpt / cpu / Forward 0.00001751 s 0.0000102036600401334 s 1.72
add_one / PartOpt / cpu / Forward 0.00001774 s 0.000010513300003367476 s 1.69
add_one / IPartOpt / cpu / Forward 0.000017356 s 0.000010385160030637054 s 1.67
add_one / DefOpt / cpu / Forward 0.000017769 s 0.000010184639986619004 s 1.74
add_one / IDefOpt / cpu / Forward 0.000017572 s 0.000010276100047121872 s 1.71
add_one / JaXPipe / cpu / PreRev 0.000019738 s 0.000011770159999286988 s 1.68
add_one / JaXPipe / cpu / PostRev 0.000019371 s 0.000011907140014955076 s 1.63
add_one / JaXPipe / cpu / BothRev 0.000019693 s 0.000012559820015667356 s 1.57
add_one / Jax / cpu / BothRev 0.000019541 s 0.00001256429996828956 s 1.56
add_one / HLOOpt / cpu / PreRev 0.000019457 s 0.000012178099977973031 s 1.60
add_one / HLOOpt / cpu / PostRev 0.000019734 s 0.00001428686002327595 s 1.38
add_one / HLOOpt / cpu / BothRev 0.000019499 s 0.00001226867996592773 s 1.59
add_one / PartOpt / cpu / PreRev 0.000019203 s 0.00001211154001794057 s 1.59
add_one / PartOpt / cpu / PostRev 0.000019776 s 0.000012167160066383077 s 1.63
add_one / PartOpt / cpu / BothRev 0.00001958 s 0.000012089080009900498 s 1.62
add_one / IPartOpt / cpu / PreRev 0.000019466 s 0.000012506240018410608 s 1.56
add_one / IPartOpt / cpu / PostRev 0.000019579 s 0.000011844299979202334 s 1.65
add_one / IPartOpt / cpu / BothRev 0.000019528 s 0.00001200290000269888 s 1.63
add_one / DefOpt / cpu / PreRev 0.000019557000000000003 s 0.000012106419962947256 s 1.62
add_one / DefOpt / cpu / PostRev 0.000019588000000000003 s 0.000011705979941325495 s 1.67
add_one / DefOpt / cpu / BothRev 0.00001989 s 0.00001240259999576665 s 1.60
add_one / IDefOpt / cpu / PreRev 0.000019556 s 0.00001211826000144356 s 1.61
add_one / IDefOpt / cpu / PostRev 0.000019594 s 0.000012116420002712402 s 1.62
add_one / IDefOpt / cpu / BothRev 0.00001972 s 0.000011627660005615326 s 1.70
add_two / JaXPipe / cpu / Primal 0.000006893759998547466 s 0.000006799080001655966 s 1.01
add_two / Jax / cpu / Primal 0.000007297300001027906 s 0.000007222659960461897 s 1.01
add_two / HLOOpt / cpu / Primal 0.000007756319998861727 s 0.000006997399977990426 s 1.11
add_two / PartOpt / cpu / Primal 0.000006786219994410204 s 0.000006764840018149698 s 1.00
add_two / IPartOpt / cpu / Primal 0.000006996700001309364 s 0.000007197699969765381 s 0.97
add_two / DefOpt / cpu / Primal 0.000006731659982506244 s 0.00000721067998711078 s 0.93
add_two / IDefOpt / cpu / Primal 0.000006849100004728825 s 0.000006686059987259796 s 1.02
add_two / JaXPipe / cpu / Forward 0.000010041079999609792 s 0.00000990931998785527 s 1.01
add_two / Jax / cpu / Forward 0.000009933800006365346 s 0.000010424320007587083 s 0.95
add_two / HLOOpt / cpu / Forward 0.000010825499996371946 s 0.000010333659993193578 s 1.05
add_two / PartOpt / cpu / Forward 0.000010270499997204752 s 0.000010246299971186093 s 1.00
add_two / IPartOpt / cpu / Forward 0.00001016749999280364 s 0.000010061119974125175 s 1.01
add_two / DefOpt / cpu / Forward 0.000010608260001845338 s 0.00001001847997031291 s 1.06
add_two / IDefOpt / cpu / Forward 0.000010061179987133071 s 0.000010443999999552032 s 0.96
add_two / JaXPipe / cpu / PreRev 0.00001396411999621705 s 0.000014581140003429028 s 0.96
add_two / JaXPipe / cpu / PostRev 0.000014164919994072989 s 0.000014217040024959715 s 1.00
add_two / JaXPipe / cpu / BothRev 0.0000141659399969285 s 0.000014091619987084412 s 1.01
add_two / Jax / cpu / BothRev 0.000013663839988566906 s 0.000014226019975467351 s 0.96
add_two / HLOOpt / cpu / PreRev 0.00001422588000423275 s 0.0000144150200412696 s 0.99
add_two / HLOOpt / cpu / PostRev 0.000016611959995316284 s 0.000018896840019806403 s 0.88
add_two / HLOOpt / cpu / BothRev 0.000013709939989894338 s 0.000013867099996787146 s 0.99
add_two / PartOpt / cpu / PreRev 0.000014172219994179614 s 0.000014706779975313113 s 0.96
add_two / PartOpt / cpu / PostRev 0.000014252340004077268 s 0.000014247260014599306 s 1.00
add_two / PartOpt / cpu / BothRev 0.00001359507999040943 s 0.000014383599991560911 s 0.95
add_two / IPartOpt / cpu / PreRev 0.00001376695999852018 s 0.00001433168001312879 s 0.96
add_two / IPartOpt / cpu / PostRev 0.00001393118000805771 s 0.000014152100011415314 s 0.98
add_two / IPartOpt / cpu / BothRev 0.000013371760001064104 s 0.000014447140029005822 s 0.93
add_two / DefOpt / cpu / PreRev 0.00001411279998592363 s 0.000014267520018620417 s 0.99
add_two / DefOpt / cpu / PostRev 0.000013884800011965126 s 0.000014259899999160552 s 0.97
add_two / DefOpt / cpu / BothRev 0.00001391903999774513 s 0.000014279120023275029 s 0.97
add_two / IDefOpt / cpu / PreRev 0.00001407152000865608 s 0.00001455288002944144 s 0.97
add_two / IDefOpt / cpu / PostRev 0.000013966780002192536 s 0.000014553979999618603 s 0.96
add_two / IDefOpt / cpu / BothRev 0.000013837540002441529 s 0.000014658500022051158 s 0.94
add_two / JaXPipe / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / Jax / cuda / Primal 0.000002432 s 0.0000019200000000000003 s 1.27
add_two / HLOOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / PartOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / IPartOpt / cuda / Primal 0.000002432 s 0.0000019200000000000003 s 1.27
add_two / DefOpt / cuda / Primal 0.000002432 s 0.0000019200000000000003 s 1.27
add_two / IDefOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / JaXPipe / cuda / Forward 0.000010495 s 0.00001008 s 1.04
add_two / Jax / cuda / Forward 0.000010432 s 0.000009824 s 1.06
add_two / HLOOpt / cuda / Forward 0.000010592 s 0.000010176 s 1.04
add_two / PartOpt / cuda / Forward 0.000010432 s 0.000009888 s 1.06
add_two / IPartOpt / cuda / Forward 0.00001024 s 0.00000944 s 1.08
add_two / DefOpt / cuda / Forward 0.000010432 s 0.000009824 s 1.06
add_two / IDefOpt / cuda / Forward 0.000010496 s 0.000010176 s 1.03
add_two / JaXPipe / cuda / PreRev 0.000032800000000000004 s 0.000032449 s 1.01
add_two / JaXPipe / cuda / PostRev 0.000032864 s 0.00003296 s 1.00
add_two / JaXPipe / cuda / BothRev 0.00003248 s 0.000032832 s 0.99
add_two / Jax / cuda / BothRev 0.000032767999999999995 s 0.000032416 s 1.01
add_two / HLOOpt / cuda / PreRev 0.000032736 s 0.000032767999999999995 s 1.00
add_two / HLOOpt / cuda / PostRev 0.000031425000000000005 s 0.000031777 s 0.99
add_two / HLOOpt / cuda / BothRev 0.000032864 s 0.000032672 s 1.01
add_two / PartOpt / cuda / PreRev 0.000033216 s 0.000032256 s 1.03
add_two / PartOpt / cuda / PostRev 0.000032736 s 0.000032160000000000004 s 1.02
add_two / PartOpt / cuda / BothRev 0.000032608 s 0.000031488 s 1.04
add_two / IPartOpt / cuda / PreRev 0.000033312 s 0.000032447 s 1.03
add_two / IPartOpt / cuda / PostRev 0.000032032 s 0.000032064 s 1.00
add_two / IPartOpt / cuda / BothRev 0.000033024 s 0.000032832 s 1.01
add_two / DefOpt / cuda / PreRev 0.000031936 s 0.000032448 s 0.98
add_two / DefOpt / cuda / PostRev 0.000032127999999999995 s 0.000032928 s 0.98
add_two / DefOpt / cuda / BothRev 0.00003232 s 0.000032672 s 0.99
add_two / IDefOpt / cuda / PreRev 0.000033503 s 0.000032384 s 1.03
add_two / IDefOpt / cuda / PostRev 0.000033376 s 0.00003296 s 1.01
add_two / IDefOpt / cuda / BothRev 0.0000336 s 0.000032 s 1.05
add_two / JaXPipe / tpu / Primal 0.000001479125 s 0.000001439975 s 1.03
add_two / Jax / tpu / Primal 0.000001427875 s 0.00000149 s 0.96
add_two / HLOOpt / tpu / Primal 0.00000148015 s 0.00000143465 s 1.03
add_two / PartOpt / tpu / Primal 0.0000014295749999999998 s 0.00000148095 s 0.97
add_two / IPartOpt / tpu / Primal 0.0000014775999999999998 s 0.000001428325 s 1.03
add_two / DefOpt / tpu / Primal 0.000001434525 s 0.00000147705 s 0.97
add_two / IDefOpt / tpu / Primal 0.00000147365 s 0.0000014395 s 1.02
add_two / JaXPipe / tpu / Forward 0.0000018161 s 0.0000018262 s 0.99
add_two / Jax / tpu / Forward 0.00000191155 s 0.0000018269 s 1.05
add_two / HLOOpt / tpu / Forward 0.0000018116 s 0.00000182515 s 0.99
add_two / PartOpt / tpu / Forward 0.000001911825 s 0.00000182465 s 1.05
add_two / IPartOpt / tpu / Forward 0.0000018185 s 0.0000018289 s 0.99
add_two / DefOpt / tpu / Forward 0.00000190855 s 0.00000183425 s 1.04
add_two / IDefOpt / tpu / Forward 0.00000183035 s 0.000001825775 s 1.00
add_two / JaXPipe / tpu / PreRev 0.0000028620250000000004 s 0.0000028433750000000005 s 1.01
add_two / JaXPipe / tpu / PostRev 0.000002720875 s 0.0000027536500000000003 s 0.99
add_two / JaXPipe / tpu / BothRev 0.000002871875 s 0.00000285435 s 1.01
add_two / Jax / tpu / BothRev 0.00000274985 s 0.0000027592 s 1.00
add_two / HLOOpt / tpu / PreRev 0.0000028666 s 0.000002848575 s 1.01
add_two / HLOOpt / tpu / PostRev 0.0000027521250000000004 s 0.0000027667000000000005 s 0.99
add_two / HLOOpt / tpu / BothRev 0.0000028813250000000003 s 0.00000284425 s 1.01
add_two / PartOpt / tpu / PreRev 0.00000275335 s 0.000002759825 s 1.00
add_two / PartOpt / tpu / PostRev 0.000002897925 s 0.000002855175 s 1.01
add_two / PartOpt / tpu / BothRev 0.000002737975 s 0.000002765525 s 0.99
add_two / IPartOpt / tpu / PreRev 0.00000287005 s 0.00000285255 s 1.01
add_two / IPartOpt / tpu / PostRev 0.00000274355 s 0.0000027584999999999995 s 0.99
add_two / IPartOpt / tpu / BothRev 0.000002880375 s 0.000002836975 s 1.02
add_two / DefOpt / tpu / PreRev 0.0000027283 s 0.00000276275 s 0.99
add_two / DefOpt / tpu / PostRev 0.0000028768749999999995 s 0.00000285745 s 1.01
add_two / DefOpt / tpu / BothRev 0.000002729525 s 0.000002763975 s 0.99
add_two / IDefOpt / tpu / PreRev 0.0000028591 s 0.0000028483000000000003 s 1.00
add_two / IDefOpt / tpu / PostRev 0.0000027312 s 0.0000027806 s 0.98
add_two / IDefOpt / tpu / BothRev 0.0000028670250000000003 s 0.000002859375 s 1.00
add_two / JaXPipe / cpu / Primal 0.000013026 s 0.000006799080001655966 s 1.92
add_two / Jax / cpu / Primal 0.000013383 s 0.000007222659960461897 s 1.85
add_two / HLOOpt / cpu / Primal 0.000013033 s 0.000006997399977990426 s 1.86
add_two / PartOpt / cpu / Primal 0.000013093 s 0.000006764840018149698 s 1.94
add_two / IPartOpt / cpu / Primal 0.000013025 s 0.000007197699969765381 s 1.81
add_two / DefOpt / cpu / Primal 0.000013309 s 0.00000721067998711078 s 1.85
add_two / IDefOpt / cpu / Primal 0.000013172 s 0.000006686059987259796 s 1.97
add_two / JaXPipe / cpu / Forward 0.000018183 s 0.00000990931998785527 s 1.83
add_two / Jax / cpu / Forward 0.000017789 s 0.000010424320007587083 s 1.71
add_two / HLOOpt / cpu / Forward 0.000017794999999999998 s 0.000010333659993193578 s 1.72
add_two / PartOpt / cpu / Forward 0.000018143 s 0.000010246299971186093 s 1.77
add_two / IPartOpt / cpu / Forward 0.000018404 s 0.000010061119974125175 s 1.83
add_two / DefOpt / cpu / Forward 0.000017988 s 0.00001001847997031291 s 1.80
add_two / IDefOpt / cpu / Forward 0.000018036 s 0.000010443999999552032 s 1.73
add_two / JaXPipe / cpu / PreRev 0.000023789 s 0.000014581140003429028 s 1.63
add_two / JaXPipe / cpu / PostRev 0.000023112 s 0.000014217040024959715 s 1.63
add_two / JaXPipe / cpu / BothRev 0.000023149 s 0.000014091619987084412 s 1.64
add_two / Jax / cpu / BothRev 0.000022507 s 0.000014226019975467351 s 1.58
add_two / HLOOpt / cpu / PreRev 0.000023484 s 0.0000144150200412696 s 1.63
add_two / HLOOpt / cpu / PostRev 0.000023149 s 0.000018896840019806403 s 1.23
add_two / HLOOpt / cpu / BothRev 0.000023364 s 0.000013867099996787146 s 1.68
add_two / PartOpt / cpu / PreRev 0.000023684 s 0.000014706779975313113 s 1.61
add_two / PartOpt / cpu / PostRev 0.000023197 s 0.000014247260014599306 s 1.63
add_two / PartOpt / cpu / BothRev 0.000022745 s 0.000014383599991560911 s 1.58
add_two / IPartOpt / cpu / PreRev 0.000023342 s 0.00001433168001312879 s 1.63
add_two / IPartOpt / cpu / PostRev 0.000022983 s 0.000014152100011415314 s 1.62
add_two / IPartOpt / cpu / BothRev 0.000022822 s 0.000014447140029005822 s 1.58
add_two / DefOpt / cpu / PreRev 0.00002337 s 0.000014267520018620417 s 1.64
add_two / DefOpt / cpu / PostRev 0.000023464 s 0.000014259899999160552 s 1.65
add_two / DefOpt / cpu / BothRev 0.000022909 s 0.000014279120023275029 s 1.60
add_two / IDefOpt / cpu / PreRev 0.000023073 s 0.00001455288002944144 s 1.59
add_two / IDefOpt / cpu / PostRev 0.000023339 s 0.000014553979999618603 s 1.60
add_two / IDefOpt / cpu / BothRev 0.000022803 s 0.000014658500022051158 s 1.56
cache / JaXPipe / cpu / Primal 0.000006618520010306384 s 0.000006307679987003212 s 1.05
cache / Jax / cpu / Primal 0.000006643239994446049 s 0.000006591660003323341 s 1.01
cache / HLOOpt / cpu / Primal 0.000006359600006362598 s 0.000006827339984738501 s 0.93
cache / PartOpt / cpu / Primal 0.000006151939996925648 s 0.000006133799961389741 s 1.00
cache / IPartOpt / cpu / Primal 0.000006265659997097828 s 0.000006438860027628834 s 0.97
cache / DefOpt / cpu / Primal 0.000006417759996111272 s 0.000006423040031222627 s 1.00
cache / IDefOpt / cpu / Primal 0.000006471060014519026 s 0.000005969559988443507 s 1.08
cache / JaXPipe / cpu / Forward 0.000016193899989502823 s 0.00001563903999340255 s 1.04
cache / Jax / cpu / Forward 0.000015538739996827645 s 0.000014838980005151826 s 1.05
cache / HLOOpt / cpu / Forward 0.00001674861999390487 s 0.000015777540002090974 s 1.06
cache / PartOpt / cpu / Forward 0.000016189159998702964 s 0.000015409440002258635 s 1.05
cache / IPartOpt / cpu / Forward 0.000016353020007500164 s 0.000015998959997887142 s 1.02
cache / DefOpt / cpu / Forward 0.000015747480001664372 s 0.000014683400013382198 s 1.07
cache / IDefOpt / cpu / Forward 0.00001579023999738638 s 0.00001531049999357492 s 1.03
cache / JaXPipe / cpu / PreRev 0.000016366140009722585 s 0.000015834379964871915 s 1.03
cache / JaXPipe / cpu / PostRev 0.000021861759985313257 s 0.00002103547999467992 s 1.04
cache / JaXPipe / cpu / BothRev 0.000017665779996605124 s 0.000017436600028304384 s 1.01
cache / Jax / cpu / BothRev 0.00002157063999447928 s 0.00003770258001168258 s 0.57
cache / HLOOpt / cpu / PreRev 0.000017247179989681172 s 0.00001725681996504136 s 1.00
cache / HLOOpt / cpu / PostRev 0.00002053518000366239 s 0.00002184574001148576 s 0.94
cache / HLOOpt / cpu / BothRev 0.00001840129999436613 s 0.00001554833997033711 s 1.18
cache / PartOpt / cpu / PreRev 0.00001763429998391075 s 0.000015944460037644604 s 1.11
cache / PartOpt / cpu / PostRev 0.000022061680008391706 s 0.00002022903999204573 s 1.09
cache / PartOpt / cpu / BothRev 0.00001771927999698164 s 0.000016357699996660814 s 1.08
cache / IPartOpt / cpu / PreRev 0.0000171317600029397 s 0.000015744579995953246 s 1.09
cache / IPartOpt / cpu / PostRev 0.000021834979993400335 s 0.00002687556004275393 s 0.81
cache / IPartOpt / cpu / BothRev 0.000016656580000926625 s 0.000016836560062074567 s 0.99
cache / DefOpt / cpu / PreRev 0.00001787870000498515 s 0.0000159212599737657 s 1.12
cache / DefOpt / cpu / PostRev 0.000017611180010135285 s 0.000016033699976105707 s 1.10
cache / DefOpt / cpu / BothRev 0.000016498019990649483 s 0.00001639453998905083 s 1.01
cache / IDefOpt / cpu / PreRev 0.000016914160005399027 s 0.00001627468000151566 s 1.04
cache / IDefOpt / cpu / PostRev 0.00001728275999539619 s 0.00001735963992359757 s 1.00
cache / IDefOpt / cpu / BothRev 0.000017963399998279783 s 0.000016383999964091344 s 1.10
cache / JaXPipe / cuda / Primal 0.000002304 s 0.000002304 s 1
cache / Jax / cuda / Primal 0.000002304 s 0.000002304 s 1
cache / HLOOpt / cuda / Primal 0.000002335 s 0.00000224 s 1.04
cache / PartOpt / cuda / Primal 0.000002335 s 0.00000224 s 1.04
cache / IPartOpt / cuda / Primal 0.000002335 s 0.000002335 s 1
cache / DefOpt / cuda / Primal 0.000002336 s 0.000002273 s 1.03
cache / IDefOpt / cuda / Primal 0.000002304 s 0.000002272 s 1.01
cache / JaXPipe / cuda / Forward 0.000002336 s 0.000002335 s 1.00
cache / Jax / cuda / Forward 0.000002336 s 0.000002335 s 1.00
cache / HLOOpt / cuda / Forward 0.000002335 s 0.000002335 s 1
cache / PartOpt / cuda / Forward 0.000002336 s 0.000002335 s 1.00
cache / IPartOpt / cuda / Forward 0.0000023670000000000004 s 0.000002335 s 1.01
cache / DefOpt / cuda / Forward 0.0000023670000000000004 s 0.000002272 s 1.04
cache / IDefOpt / cuda / Forward 0.000002336 s 0.000002336 s 1
cache / JaXPipe / cuda / PreRev 0.000010624 s 0.00001088 s 0.98
cache / JaXPipe / cuda / PostRev 0.000010688 s 0.000011008 s 0.97
cache / JaXPipe / cuda / BothRev 0.000010624 s 0.000010591 s 1.00
cache / Jax / cuda / BothRev 0.00001072 s 0.000010975 s 0.98
cache / HLOOpt / cuda / PreRev 0.000013727 s 0.000013408 s 1.02
cache / HLOOpt / cuda / PostRev 0.000013664 s 0.000013409000000000002 s 1.02
cache / HLOOpt / cuda / BothRev 0.000013729 s 0.000013408 s 1.02
cache / PartOpt / cuda / PreRev 0.000010816 s 0.00001088 s 0.99
cache / PartOpt / cuda / PostRev 0.000010592 s 0.00001104 s 0.96
cache / PartOpt / cuda / BothRev 0.0000104 s 0.000011007 s 0.94
cache / IPartOpt / cuda / PreRev 0.000010272 s 0.000010688 s 0.96
cache / IPartOpt / cuda / PostRev 0.000011007 s 0.000010944 s 1.01
cache / IPartOpt / cuda / BothRev 0.000010688 s 0.000010912 s 0.98
cache / DefOpt / cuda / PreRev 0.000010656 s 0.000010816 s 0.99
cache / DefOpt / cuda / PostRev 0.000010784 s 0.000010784 s 1
cache / DefOpt / cuda / BothRev 0.00001088 s 0.000010785 s 1.01
cache / IDefOpt / cuda / PreRev 0.000010591 s 0.000010912 s 0.97
cache / IDefOpt / cuda / PostRev 0.000010912 s 0.00001088 s 1.00
cache / IDefOpt / cuda / BothRev 0.000010337 s 0.000010688 s 0.97
cache / JaXPipe / tpu / Primal 0.000002459225 s 0.00000245085 s 1.00
cache / Jax / tpu / Primal 0.00000244925 s 0.0000024729 s 0.99
cache / HLOOpt / tpu / Primal 0.00000246115 s 0.0000024556500000000004 s 1.00
cache / PartOpt / tpu / Primal 0.000002456525 s 0.0000024559 s 1.00
cache / IPartOpt / tpu / Primal 0.0000024783 s 0.0000024524 s 1.01
cache / DefOpt / tpu / Primal 0.00000246685 s 0.0000024672000000000003 s 1.00
cache / IDefOpt / tpu / Primal 0.00000246785 s 0.000002476925 s 1.00
cache / JaXPipe / tpu / Forward 0.000003563275 s 0.0000035351500000000004 s 1.01
cache / Jax / tpu / Forward 0.000003555475 s 0.0000035542 s 1.00
cache / HLOOpt / tpu / Forward 0.000003568375 s 0.00000355125 s 1.00
cache / PartOpt / tpu / Forward 0.000003555275 s 0.000003535775 s 1.01
cache / IPartOpt / tpu / Forward 0.000003582375 s 0.0000035647250000000004 s 1.00
cache / DefOpt / tpu / Forward 0.0000035395500000000003 s 0.000003530025 s 1.00
cache / IDefOpt / tpu / Forward 0.0000035557 s 0.00000355895 s 1.00
cache / JaXPipe / tpu / PreRev 0.000004971425 s 0.000004948925 s 1.00
cache / JaXPipe / tpu / PostRev 0.000004988 s 0.000004943575 s 1.01
cache / JaXPipe / tpu / BothRev 0.000005011025 s 0.0000049926 s 1.00
cache / Jax / tpu / BothRev 0.000005000625 s 0.0000049899 s 1.00
cache / HLOOpt / tpu / PreRev 0.000003964050000000001 s 0.000003952875 s 1.00
cache / HLOOpt / tpu / PostRev 0.0000041734 s 0.000004132775 s 1.01
cache / HLOOpt / tpu / BothRev 0.00000397095 s 0.000003959050000000001 s 1.00
cache / PartOpt / tpu / PreRev 0.000005014475000000001 s 0.0000049826 s 1.01
cache / PartOpt / tpu / PostRev 0.0000050026 s 0.00000497455 s 1.01
cache / PartOpt / tpu / BothRev 0.000005010425 s 0.00000497915 s 1.01
cache / IPartOpt / tpu / PreRev 0.000005008399999999999 s 0.0000049774 s 1.01
cache / IPartOpt / tpu / PostRev 0.000005029875000000001 s 0.0000049795500000000005 s 1.01
cache / IPartOpt / tpu / BothRev 0.0000049873 s 0.000004968025 s 1.00
cache / DefOpt / tpu / PreRev 0.0000049997 s 0.0000049699000000000005 s 1.01
cache / DefOpt / tpu / PostRev 0.000004989025 s 0.0000049801 s 1.00
cache / DefOpt / tpu / BothRev 0.000005009375 s 0.00000495775 s 1.01
cache / IDefOpt / tpu / PreRev 0.00000499095 s 0.000004973275 s 1.00
cache / IDefOpt / tpu / PostRev 0.00000501475 s 0.000004974475 s 1.01
cache / IDefOpt / tpu / BothRev 0.000004983125 s 0.000004971775 s 1.00
cache / JaXPipe / cpu / Primal 0.000012733 s 0.000006307679987003212 s 2.02
cache / Jax / cpu / Primal 0.000012728 s 0.000006591660003323341 s 1.93
cache / HLOOpt / cpu / Primal 0.000012774 s 0.000006827339984738501 s 1.87
cache / PartOpt / cpu / Primal 0.000012842 s 0.000006133799961389741 s 2.09
cache / IPartOpt / cpu / Primal 0.000012907 s 0.000006438860027628834 s 2.00
cache / DefOpt / cpu / Primal 0.000012712 s 0.000006423040031222627 s 1.98
cache / IDefOpt / cpu / Primal 0.000012283 s 0.000005969559988443507 s 2.06
cache / JaXPipe / cpu / Forward 0.000017117 s 0.00001563903999340255 s 1.09
cache / Jax / cpu / Forward 0.000017086000000000002 s 0.000014838980005151826 s 1.15
cache / HLOOpt / cpu / Forward 0.000017153 s 0.000015777540002090974 s 1.09
cache / PartOpt / cpu / Forward 0.000016765 s 0.000015409440002258635 s 1.09
cache / IPartOpt / cpu / Forward 0.000027259 s 0.000015998959997887142 s 1.70
cache / DefOpt / cpu / Forward 0.000016995 s 0.000014683400013382198 s 1.16
cache / IDefOpt / cpu / Forward 0.000017142 s 0.00001531049999357492 s 1.12
cache / JaXPipe / cpu / PreRev 0.000017156 s 0.000015834379964871915 s 1.08
cache / JaXPipe / cpu / PostRev 0.000019799 s 0.00002103547999467992 s 0.94
cache / JaXPipe / cpu / BothRev 0.00001763 s 0.000017436600028304384 s 1.01
cache / Jax / cpu / BothRev 0.000020199 s 0.00003770258001168258 s 0.54
cache / HLOOpt / cpu / PreRev 0.000017298 s 0.00001725681996504136 s 1.00
cache / HLOOpt / cpu / PostRev 0.000017772 s 0.00002184574001148576 s 0.81
cache / HLOOpt / cpu / BothRev 0.000017215 s 0.00001554833997033711 s 1.11
cache / PartOpt / cpu / PreRev 0.000017718999999999998 s 0.000015944460037644604 s 1.11
cache / PartOpt / cpu / PostRev 0.000019934 s 0.00002022903999204573 s 0.99
cache / PartOpt / cpu / BothRev 0.00001764 s 0.000016357699996660814 s 1.08
cache / IPartOpt / cpu / PreRev 0.000017773 s 0.000015744579995953246 s 1.13
cache / IPartOpt / cpu / PostRev 0.000020288 s 0.00002687556004275393 s 0.75
cache / IPartOpt / cpu / BothRev 0.000017431 s 0.000016836560062074567 s 1.04
cache / DefOpt / cpu / PreRev 0.000017308 s 0.0000159212599737657 s 1.09
cache / DefOpt / cpu / PostRev 0.000017411 s 0.000016033699976105707 s 1.09
cache / DefOpt / cpu / BothRev 0.000017592 s 0.00001639453998905083 s 1.07
cache / IDefOpt / cpu / PreRev 0.000017743999999999998 s 0.00001627468000151566 s 1.09
cache / IDefOpt / cpu / PostRev 0.000017639 s 0.00001735963992359757 s 1.02
cache / IDefOpt / cpu / BothRev 0.00001788 s 0.000016383999964091344 s 1.09
Concat / JaXPipe / cpu / Primal 0.000006840439993993641 s 0.000007008039992797421 s 0.98
Concat / Jax / cpu / Primal 0.000006571220001205802 s 0.000006910560005053412 s 0.95
Concat / HLOOpt / cpu / Primal 0.000007116079996194458 s 0.0000066355800026940416 s 1.07
Concat / PartOpt / cpu / Primal 0.000007026359996871179 s 0.000006315100017673104 s 1.11
Concat / IPartOpt / cpu / Primal 0.000007231519996366842 s 0.00000672385998768732 s 1.08
Concat / DefOpt / cpu / Primal 0.000006638859997565305 s 0.000006603359997825464 s 1.01
Concat / IDefOpt / cpu / Primal 0.00000688508000166621 s 0.000006776739974156954 s 1.02
Concat / JaXPipe / cpu / Forward 0.000010559339993960748 s 0.00000970459999734885 s 1.09
Concat / Jax / cpu / Forward 0.000010335900003610732 s 0.000009664840008554166 s 1.07
Concat / HLOOpt / cpu / Forward 0.000010383799999544864 s 0.000009945999991032295 s 1.04
Concat / PartOpt / cpu / Forward 0.000010360140004195272 s 0.000010298160013917368 s 1.01
Concat / IPartOpt / cpu / Forward 0.00000998173999505525 s 0.000010229179988527903 s 0.98
Concat / DefOpt / cpu / Forward 0.00001016587999629337 s 0.000009792959981496096 s 1.04
Concat / IDefOpt / cpu / Forward 0.000010304060006092184 s 0.000009998619980251533 s 1.03
Concat / JaXPipe / cpu / PreRev 0.000012129780000122991 s 0.00001109117998566944 s 1.09
Concat / JaXPipe / cpu / PostRev 0.00001170631999229954 s 0.000011185500006831715 s 1.05
Concat / JaXPipe / cpu / BothRev 0.000011761360001401044 s 0.000011662339984468415 s 1.01
Concat / Jax / cpu / BothRev 0.000012005260000478302 s 0.000011574699983611936 s 1.04
Concat / HLOOpt / cpu / PreRev 0.000012441160004073028 s 0.000011854180002046633 s 1.05
Concat / HLOOpt / cpu / PostRev 0.000013901080005780388 s 0.000013122439995640888 s 1.06
Concat / HLOOpt / cpu / BothRev 0.000011504680005600675 s 0.000011466600044514053 s 1.00
Concat / PartOpt / cpu / PreRev 0.000011865980004586163 s 0.000011967220025326242 s 0.99
Concat / PartOpt / cpu / PostRev 0.000011784399994212436 s 0.00001147991998550424 s 1.03
Concat / PartOpt / cpu / BothRev 0.000012109659994621325 s 0.000012181980009700055 s 0.99
Concat / IPartOpt / cpu / PreRev 0.000011801199998444644 s 0.000012034939991281136 s 0.98
Concat / IPartOpt / cpu / PostRev 0.000011645139993561315 s 0.000011964140003328794 s 0.97
Concat / IPartOpt / cpu / BothRev 0.000011904199998298282 s 0.000011217660012334816 s 1.06
Concat / DefOpt / cpu / PreRev 0.000012290919994484283 s 0.000011659739957394775 s 1.05
Concat / DefOpt / cpu / PostRev 0.000011664379997000651 s 0.000011703559994202806 s 1.00
Concat / DefOpt / cpu / BothRev 0.00001192483999830074 s 0.0000116978799997014 s 1.02
Concat / IDefOpt / cpu / PreRev 0.000011999760001799586 s 0.00001161827999567322 s 1.03
Concat / IDefOpt / cpu / PostRev 0.000011523079999733454 s 0.000011777499994423124 s 0.98
Concat / IDefOpt / cpu / BothRev 0.000011586499992972676 s 0.000011760860015783692 s 0.99
Concat / JaXPipe / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / Jax / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / HLOOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / PartOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / IPartOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / DefOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / IDefOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / JaXPipe / cuda / Forward 0.000010944 s 0.000009824 s 1.11
Concat / Jax / cuda / Forward 0.000010816 s 0.000009568 s 1.13
Concat / HLOOpt / cuda / Forward 0.00001056 s 0.000010144 s 1.04
Concat / PartOpt / cuda / Forward 0.000010816 s 0.000009856 s 1.10
Concat / IPartOpt / cuda / Forward 0.000010912 s 0.000010368 s 1.05
Concat / DefOpt / cuda / Forward 0.000010783 s 0.00000992 s 1.09
Concat / IDefOpt / cuda / Forward 0.000010656 s 0.000009888 s 1.08
Concat / JaXPipe / cuda / PreRev 0.000016831 s 0.000016768000000000003 s 1.00
Concat / JaXPipe / cuda / PostRev 0.000016705 s 0.000016831 s 0.99
Concat / JaXPipe / cuda / BothRev 0.000016833 s 0.00001632 s 1.03
Concat / Jax / cuda / BothRev 0.000017184 s 0.000016737 s 1.03
Concat / HLOOpt / cuda / PreRev 0.000016512 s 0.000016608 s 0.99
Concat / HLOOpt / cuda / PostRev 0.000017216 s 0.00001632 s 1.05
Concat / HLOOpt / cuda / BothRev 0.000017184 s 0.00001616 s 1.06
Concat / PartOpt / cuda / PreRev 0.000017184 s 0.00001712 s 1.00
Concat / PartOpt / cuda / PostRev 0.000017344 s 0.000016670999999999997 s 1.04
Concat / PartOpt / cuda / BothRev 0.000016768000000000003 s 0.00001696 s 0.99
Concat / IPartOpt / cuda / PreRev 0.000017184 s 0.000017152 s 1.00
Concat / IPartOpt / cuda / PostRev 0.00001648 s 0.000016383999999999998 s 1.01
Concat / IPartOpt / cuda / BothRev 0.000016768000000000003 s 0.000016736 s 1.00
Concat / DefOpt / cuda / PreRev 0.000016672 s 0.000016416 s 1.02
Concat / DefOpt / cuda / PostRev 0.00001664 s 0.000016927999999999998 s 0.98
Concat / DefOpt / cuda / BothRev 0.000016991 s 0.000016927999999999998 s 1.00
Concat / IDefOpt / cuda / PreRev 0.00001696 s 0.000024671 s 0.69
Concat / IDefOpt / cuda / PostRev 0.000017375999999999998 s 0.000016703 s 1.04
Concat / IDefOpt / cuda / BothRev 0.000025663 s 0.00001648 s 1.56
Concat / JaXPipe / tpu / Primal 0.000001529075 s 0.000001524175 s 1.00
Concat / Jax / tpu / Primal 0.0000015675749999999998 s 0.0000015307999999999998 s 1.02
Concat / HLOOpt / tpu / Primal 0.0000015338999999999998 s 0.000001524975 s 1.01
Concat / PartOpt / tpu / Primal 0.0000015602250000000002 s 0.00000153215 s 1.02
Concat / IPartOpt / tpu / Primal 0.000001526975 s 0.00000152025 s 1.00
Concat / DefOpt / tpu / Primal 0.0000015676000000000002 s 0.000001536925 s 1.02
Concat / IDefOpt / tpu / Primal 0.0000015304000000000002 s 0.000001526325 s 1.00
Concat / JaXPipe / tpu / Forward 0.0000015769250000000002 s 0.0000015745 s 1.00
Concat / Jax / tpu / Forward 0.0000015600249999999995 s 0.0000015418 s 1.01
Concat / HLOOpt / tpu / Forward 0.0000015849750000000002 s 0.000001583675 s 1.00
Concat / PartOpt / tpu / Forward 0.000001564875 s 0.0000015417 s 1.02
Concat / IPartOpt / tpu / Forward 0.00000156205 s 0.00000158935 s 0.98
Concat / DefOpt / tpu / Forward 0.00000157275 s 0.000001551225 s 1.01
Concat / IDefOpt / tpu / Forward 0.0000015602250000000002 s 0.0000015831500000000002 s 0.99
Concat / JaXPipe / tpu / PreRev 0.0000019973 s 0.0000019998 s 1.00
Concat / JaXPipe / tpu / PostRev 0.00000204235 s 0.0000020947 s 0.98
Concat / JaXPipe / tpu / BothRev 0.00000199965 s 0.000002003425 s 1.00
Concat / Jax / tpu / BothRev 0.000002046625 s 0.000002079375 s 0.98
Concat / HLOOpt / tpu / PreRev 0.0000019987 s 0.0000019996249999999995 s 1.00
Concat / HLOOpt / tpu / PostRev 0.00000204295 s 0.000002077325 s 0.98
Concat / HLOOpt / tpu / BothRev 0.000002008475 s 0.0000019949 s 1.01
Concat / PartOpt / tpu / PreRev 0.000002046075 s 0.00000207675 s 0.99
Concat / PartOpt / tpu / PostRev 0.0000019998250000000004 s 0.000001998025 s 1.00
Concat / PartOpt / tpu / BothRev 0.000002051325 s 0.0000020791 s 0.99
Concat / IPartOpt / tpu / PreRev 0.0000020175750000000003 s 0.000001992975 s 1.01
Concat / IPartOpt / tpu / PostRev 0.000002051075 s 0.000002076225 s 0.99
Concat / IPartOpt / tpu / BothRev 0.000002000275 s 0.0000019937 s 1.00
Concat / DefOpt / tpu / PreRev 0.0000020408 s 0.000002078075 s 0.98
Concat / DefOpt / tpu / PostRev 0.00000200235 s 0.00000199665 s 1.00
Concat / DefOpt / tpu / BothRev 0.00000204865 s 0.00000208455 s 0.98
Concat / IDefOpt / tpu / PreRev 0.0000020038 s 0.0000020115 s 1.00
Concat / IDefOpt / tpu / PostRev 0.0000020599 s 0.00000208505 s 0.99
Concat / IDefOpt / tpu / BothRev 0.000002002225 s 0.0000020053750000000003 s 1.00
Concat / JaXPipe / cpu / Primal 0.000013058 s 0.000007008039992797421 s 1.86
Concat / Jax / cpu / Primal 0.000012917 s 0.000006910560005053412 s 1.87
Concat / HLOOpt / cpu / Primal 0.000012803 s 0.0000066355800026940416 s 1.93
Concat / PartOpt / cpu / Primal 0.000012577 s 0.000006315100017673104 s 1.99
Concat / IPartOpt / cpu / Primal 0.000012648 s 0.00000672385998768732 s 1.88
Concat / DefOpt / cpu / Primal 0.000012625 s 0.000006603359997825464 s 1.91
Concat / IDefOpt / cpu / Primal 0.000012503 s 0.000006776739974156954 s 1.84
Concat / JaXPipe / cpu / Forward 0.000017636 s 0.00000970459999734885 s 1.82
Concat / Jax / cpu / Forward 0.000017519 s 0.000009664840008554166 s 1.81
Concat / HLOOpt / cpu / Forward 0.000017239999999999998 s 0.000009945999991032295 s 1.73
Concat / PartOpt / cpu / Forward 0.000017685 s 0.000010298160013917368 s 1.72
Concat / IPartOpt / cpu / Forward 0.0000176 s 0.000010229179988527903 s 1.72
Concat / DefOpt / cpu / Forward 0.000017197 s 0.000009792959981496096 s 1.76
Concat / IDefOpt / cpu / Forward 0.000017373 s 0.000009998619980251533 s 1.74
Concat / JaXPipe / cpu / PreRev 0.000020518 s 0.00001109117998566944 s 1.85
Concat / JaXPipe / cpu / PostRev 0.000019712 s 0.000011185500006831715 s 1.76
Concat / JaXPipe / cpu / BothRev 0.000019548 s 0.000011662339984468415 s 1.68
Concat / Jax / cpu / BothRev 0.000019955 s 0.000011574699983611936 s 1.72
Concat / HLOOpt / cpu / PreRev 0.000019978 s 0.000011854180002046633 s 1.69
Concat / HLOOpt / cpu / PostRev 0.000019835 s 0.000013122439995640888 s 1.51
Concat / HLOOpt / cpu / BothRev 0.000019794 s 0.000011466600044514053 s 1.73
Concat / PartOpt / cpu / PreRev 0.000020240000000000003 s 0.000011967220025326242 s 1.69
Concat / PartOpt / cpu / PostRev 0.000020082 s 0.00001147991998550424 s 1.75
Concat / PartOpt / cpu / BothRev 0.000019621 s 0.000012181980009700055 s 1.61
Concat / IPartOpt / cpu / PreRev 0.000019987 s 0.000012034939991281136 s 1.66
Concat / IPartOpt / cpu / PostRev 0.000019651 s 0.000011964140003328794 s 1.64
Concat / IPartOpt / cpu / BothRev 0.000019521 s 0.000011217660012334816 s 1.74
Concat / DefOpt / cpu / PreRev 0.000019755000000000003 s 0.000011659739957394775 s 1.69
Concat / DefOpt / cpu / PostRev 0.000019978 s 0.000011703559994202806 s 1.71
Concat / DefOpt / cpu / BothRev 0.000019586 s 0.0000116978799997014 s 1.67
Concat / IDefOpt / cpu / PreRev 0.000019666 s 0.00001161827999567322 s 1.69
Concat / IDefOpt / cpu / PostRev 0.00001974 s 0.000011777499994423124 s 1.68
Concat / IDefOpt / cpu / BothRev 0.000019385 s 0.000011760860015783692 s 1.65
const_scatter / JaXPipe / cpu / Primal 0.000006339839992506313 s 0.0000062349000108952165 s 1.02
const_scatter / Jax / cpu / Primal 0.000006830140002875851 s 0.00000655261999781942 s 1.04
const_scatter / HLOOpt / cpu / Primal 0.0000072999400072149 s 0.000007446100034940173 s 0.98
const_scatter / PartOpt / cpu / Primal 0.0000062977199900160485 s 0.000006509699978778372 s 0.97
const_scatter / IPartOpt / cpu / Primal 0.000006772659994567221 s 0.000006898999972690945 s 0.98
const_scatter / DefOpt / cpu / Primal 0.000006847839999863936 s 0.0000069046599946887 s 0.99
const_scatter / IDefOpt / cpu / Primal 0.000006797580003876646 s 0.00000668348001454433 s 1.02
const_scatter / JaXPipe / cpu / Forward 0.000010918739990302129 s 0.00001064506002876442 s 1.03
const_scatter / Jax / cpu / Forward 0.00000955431999727807 s 0.00000971563999883074 s 0.98
const_scatter / HLOOpt / cpu / Forward 0.000011318060016947128 s 0.00001086616000975482 s 1.04
const_scatter / PartOpt / cpu / Forward 0.00001049362000230758 s 0.00001097888003641856 s 0.96
const_scatter / IPartOpt / cpu / Forward 0.000011357259991200408 s 0.000011530499987202348 s 0.98
const_scatter / DefOpt / cpu / Forward 0.000010932260008758022 s 0.00001086903999748756 s 1.01
const_scatter / IDefOpt / cpu / Forward 0.000010383020012341148 s 0.00001085090003471123 s 0.96
const_scatter / JaXPipe / cpu / PreRev 0.0002849162199959 s 0.0003115300599802 s 0.91
const_scatter / JaXPipe / cpu / PostRev 0.0002798873400001 s 0.0002813080199848 s 0.99
const_scatter / JaXPipe / cpu / BothRev 0.0002833825999891 s 0.0002825666399712 s 1.00
const_scatter / Jax / cpu / BothRev 0.0002834577400108 s 0.0002801638399796 s 1.01
const_scatter / HLOOpt / cpu / PreRev 0.0002851773199927 s 0.0002834601199992 s 1.01
const_scatter / HLOOpt / cpu / PostRev 0.0002843868199943 s 0.0002848288000222 s 1.00
const_scatter / HLOOpt / cpu / BothRev 0.0002856843000017 s 0.0002825747000315 s 1.01
const_scatter / PartOpt / cpu / PreRev 0.0002830665999908 s 0.0002811166400533 s 1.01
const_scatter / PartOpt / cpu / PostRev 0.0002808964799987 s 0.0002791229000013 s 1.01
const_scatter / PartOpt / cpu / BothRev 0.0002824971600034 s 0.000281865579991 s 1.00
const_scatter / IPartOpt / cpu / PreRev 0.000283128660003 s 0.0002801982199798 s 1.01
const_scatter / IPartOpt / cpu / PostRev 0.0002842843599933 s 0.0002806498799873 s 1.01
const_scatter / IPartOpt / cpu / BothRev 0.0002909773400028 s 0.0002829532800205 s 1.03
const_scatter / DefOpt / cpu / PreRev 0.0002833901400026 s 0.0002809657799934 s 1.01
const_scatter / DefOpt / cpu / PostRev 0.0002822115800017 s 0.0002825006799594 s 1.00
const_scatter / DefOpt / cpu / BothRev 0.0002851361199918 s 0.0002812875200197 s 1.01
const_scatter / IDefOpt / cpu / PreRev 0.0002829455200048 s 0.0002909929399902 s 0.97
const_scatter / IDefOpt / cpu / PostRev 0.0002831455200021 s 0.0002834680999694 s 1.00
const_scatter / IDefOpt / cpu / BothRev 0.000285003580002 s 0.0002822982999714 s 1.01
const_scatter / JaXPipe / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / Jax / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / HLOOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / PartOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / IPartOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / DefOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / IDefOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / JaXPipe / cuda / Forward 0.000010529 s 0.000010113 s 1.04
const_scatter / Jax / cuda / Forward 0.000010464 s 0.000009824 s 1.07
const_scatter / HLOOpt / cuda / Forward 0.000010336 s 0.00000992 s 1.04
const_scatter / PartOpt / cuda / Forward 0.000010592 s 0.000009984 s 1.06
const_scatter / IPartOpt / cuda / Forward 0.000012224 s 0.00000992 s 1.23
const_scatter / DefOpt / cuda / Forward 0.000010464 s 0.000009856 s 1.06
const_scatter / IDefOpt / cuda / Forward 0.000010496 s 0.000009504 s 1.10
const_scatter / JaXPipe / cuda / PreRev 0.000016127 s 0.000016576000000000002 s 0.97
const_scatter / JaXPipe / cuda / PostRev 0.000016608 s 0.00001584 s 1.05
const_scatter / JaXPipe / cuda / BothRev 0.000016416 s 0.000016705 s 0.98
const_scatter / Jax / cuda / BothRev 0.00002272 s 0.000016864 s 1.35
const_scatter / HLOOpt / cuda / PreRev 0.000016896000000000002 s 0.000016737 s 1.01
const_scatter / HLOOpt / cuda / PostRev 0.000016832 s 0.00001712 s 0.98
const_scatter / HLOOpt / cuda / BothRev 0.000016544 s 0.000016385 s 1.01
const_scatter / PartOpt / cuda / PreRev 0.000016576000000000002 s 0.000016832 s 0.98
const_scatter / PartOpt / cuda / PostRev 0.00001648 s 0.00001568 s 1.05
const_scatter / PartOpt / cuda / BothRev 0.000016832 s 0.000015937 s 1.06
const_scatter / IPartOpt / cuda / PreRev 0.00001664 s 0.000016768999999999998 s 0.99
const_scatter / IPartOpt / cuda / PostRev 0.000016353 s 0.000017984 s 0.91
const_scatter / IPartOpt / cuda / BothRev 0.000016832 s 0.000016096 s 1.05
const_scatter / DefOpt / cuda / PreRev 0.000016192 s 0.000016736 s 0.97
const_scatter / DefOpt / cuda / PostRev 0.000016608 s 0.000016576000000000002 s 1.00
const_scatter / DefOpt / cuda / BothRev 0.000016735 s 0.000017312 s 0.97
const_scatter / IDefOpt / cuda / PreRev 0.0000168 s 0.000016705 s 1.01
const_scatter / IDefOpt / cuda / PostRev 0.000016864 s 0.000016608 s 1.02
const_scatter / IDefOpt / cuda / BothRev 0.000016608 s 0.000016257 s 1.02
const_scatter / JaXPipe / tpu / Primal 0.000003835425 s 0.000003827175 s 1.00
const_scatter / Jax / tpu / Primal 0.000003802675 s 0.000003828075000000001 s 0.99
const_scatter / HLOOpt / tpu / Primal 0.0000038187500000000005 s 0.00000381715 s 1.00
const_scatter / PartOpt / tpu / Primal 0.00000379185 s 0.00000383065 s 0.99
const_scatter / IPartOpt / tpu / Primal 0.0000038365500000000006 s 0.0000038145 s 1.01
const_scatter / DefOpt / tpu / Primal 0.0000037936 s 0.000003827675 s 0.99
const_scatter / IDefOpt / tpu / Primal 0.0000038550500000000005 s 0.00000379325 s 1.02
const_scatter / JaXPipe / tpu / Forward 0.000006468325 s 0.000006433525 s 1.01
const_scatter / Jax / tpu / Forward 0.000006474175 s 0.000006517650000000001 s 0.99
const_scatter / HLOOpt / tpu / Forward 0.0000064729 s 0.0000064666 s 1.00
const_scatter / PartOpt / tpu / Forward 0.00000646595 s 0.0000065188 s 0.99
const_scatter / IPartOpt / tpu / Forward 0.000006465625 s 0.000006463575 s 1.00
const_scatter / DefOpt / tpu / Forward 0.000006479225 s 0.0000065058000000000005 s 1.00
const_scatter / IDefOpt / tpu / Forward 0.000006473149999999999 s 0.0000064621 s 1.00
const_scatter / JaXPipe / tpu / PreRev 0.000006675675 s 0.000006682225000000001 s 1.00
const_scatter / JaXPipe / tpu / PostRev 0.000006650899999999999 s 0.0000066495 s 1.00
const_scatter / JaXPipe / tpu / BothRev 0.0000066646000000000005 s 0.000006660075 s 1.00
const_scatter / Jax / tpu / BothRev 0.000006659399999999999 s 0.000006669049999999999 s 1.00
const_scatter / HLOOpt / tpu / PreRev 0.0000066895 s 0.0000066533750000000005 s 1.01
const_scatter / HLOOpt / tpu / PostRev 0.00000665055 s 0.000006653775 s 1.00
const_scatter / HLOOpt / tpu / BothRev 0.0000066727250000000006 s 0.000006698075 s 1.00
const_scatter / PartOpt / tpu / PreRev 0.000006652275 s 0.000006674774999999999 s 1.00
const_scatter / PartOpt / tpu / PostRev 0.00000668295 s 0.000006659625 s 1.00
const_scatter / PartOpt / tpu / BothRev 0.0000066414 s 0.0000066755 s 0.99
const_scatter / IPartOpt / tpu / PreRev 0.0000066881500000000005 s 0.000006695950000000001 s 1.00
const_scatter / IPartOpt / tpu / PostRev 0.000006637875 s 0.000006663375 s 1.00
const_scatter / IPartOpt / tpu / BothRev 0.000006655474999999999 s 0.000006657750000000001 s 1.00
const_scatter / DefOpt / tpu / PreRev 0.0000066599750000000006 s 0.00000665115 s 1.00
const_scatter / DefOpt / tpu / PostRev 0.00000665435 s 0.0000066825 s 1.00
const_scatter / DefOpt / tpu / BothRev 0.00000664515 s 0.000006679575 s 0.99
const_scatter / IDefOpt / tpu / PreRev 0.0000066816 s 0.000006672949999999999 s 1.00
const_scatter / IDefOpt / tpu / PostRev 0.0000066425500000000005 s 0.0000066757750000000005 s 1.00
const_scatter / IDefOpt / tpu / BothRev 0.0000066682 s 0.00000667055 s 1.00
const_scatter / JaXPipe / cpu / Primal 0.000012853 s 0.0000062349000108952165 s 2.06
const_scatter / Jax / cpu / Primal 0.000012558 s 0.00000655261999781942 s 1.92
const_scatter / HLOOpt / cpu / Primal 0.000013363 s 0.000007446100034940173 s 1.79
const_scatter / PartOpt / cpu / Primal 0.000012692 s 0.000006509699978778372 s 1.95
const_scatter / IPartOpt / cpu / Primal 0.000012613 s 0.000006898999972690945 s 1.83
const_scatter / DefOpt / cpu / Primal 0.000013262 s 0.0000069046599946887 s 1.92
const_scatter / IDefOpt / cpu / Primal 0.00001332 s 0.00000668348001454433 s 1.99
const_scatter / JaXPipe / cpu / Forward 0.000018098 s 0.00001064506002876442 s 1.70
const_scatter / Jax / cpu / Forward 0.000016714 s 0.00000971563999883074 s 1.72
const_scatter / HLOOpt / cpu / Forward 0.000017624 s 0.00001086616000975482 s 1.62
const_scatter / PartOpt / cpu / Forward 0.00001747 s 0.00001097888003641856 s 1.59
const_scatter / IPartOpt / cpu / Forward 0.000017447 s 0.000011530499987202348 s 1.51
const_scatter / DefOpt / cpu / Forward 0.000017894000000000002 s 0.00001086903999748756 s 1.65
const_scatter / IDefOpt / cpu / Forward 0.000017989 s 0.00001085090003471123 s 1.66
const_scatter / JaXPipe / cpu / PreRev 0.000495964 s 0.0003115300599802 s 1.59
const_scatter / JaXPipe / cpu / PostRev 0.000490749 s 0.0002813080199848 s 1.74
const_scatter / JaXPipe / cpu / BothRev 0.0004996789999999 s 0.0002825666399712 s 1.77
const_scatter / Jax / cpu / BothRev 0.000488794 s 0.0002801638399796 s 1.74
const_scatter / HLOOpt / cpu / PreRev 0.000509893 s 0.0002834601199992 s 1.80
const_scatter / HLOOpt / cpu / PostRev 0.000506054 s 0.0002848288000222 s 1.78
const_scatter / HLOOpt / cpu / BothRev 0.000499043 s 0.0002825747000315 s 1.77
const_scatter / PartOpt / cpu / PreRev 0.000487932 s 0.0002811166400533 s 1.74
const_scatter / PartOpt / cpu / PostRev 0.000505576 s 0.0002791229000013 s 1.81
const_scatter / PartOpt / cpu / BothRev 0.000516903 s 0.000281865579991 s 1.83
const_scatter / IPartOpt / cpu / PreRev 0.000490302 s 0.0002801982199798 s 1.75
const_scatter / IPartOpt / cpu / PostRev 0.000511907 s 0.0002806498799873 s 1.82
const_scatter / IPartOpt / cpu / BothRev 0.000508716 s 0.0002829532800205 s 1.80
const_scatter / DefOpt / cpu / PreRev 0.000512936 s 0.0002809657799934 s 1.83
const_scatter / DefOpt / cpu / PostRev 0.000507228 s 0.0002825006799594 s 1.80
const_scatter / DefOpt / cpu / BothRev 0.000506519 s 0.0002812875200197 s 1.80
const_scatter / IDefOpt / cpu / PreRev 0.000508772 s 0.0002909929399902 s 1.75
const_scatter / IDefOpt / cpu / PostRev 0.000489378 s 0.0002834680999694 s 1.73
const_scatter / IDefOpt / cpu / BothRev 0.000513575 s 0.0002822982999714 s 1.82
GenDot / JaXPipe / cpu / Primal 0.000007136959995932557 s 0.000007543960009570583 s 0.95
GenDot / Jax / cpu / Primal 0.000007347880000452278 s 0.00000684336000631447 s 1.07
GenDot / HLOOpt / cpu / Primal 0.000007859139993797725 s 0.000007792519991198787 s 1.01
GenDot / PartOpt / cpu / Primal 0.000006978460003210785 s 0.000006615960010094568 s 1.05
GenDot / IPartOpt / cpu / Primal 0.000007079060001160542 s 0.000006698160013911547 s 1.06
GenDot / DefOpt / cpu / Primal 0.000007359299991094303 s 0.000007053299959807191 s 1.04
GenDot / IDefOpt / cpu / Primal 0.000007659480002075724 s 0.00000699161998454656 s 1.10
GenDot / JaXPipe / cpu / Forward 0.000010983559998294369 s 0.000010652839973772645 s 1.03
GenDot / Jax / cpu / Forward 0.000010310600011962378 s 0.000010071080032503232 s 1.02
GenDot / HLOOpt / cpu / Forward 0.000011651739996523248 s 0.000011345020020598896 s 1.03
GenDot / PartOpt / cpu / Forward 0.000010776600011013217 s 0.000010618520036587142 s 1.01
GenDot / IPartOpt / cpu / Forward 0.00001200288000063665 s 0.000011141940003653872 s 1.08
GenDot / DefOpt / cpu / Forward 0.000011129279992019292 s 0.000011015899963240372 s 1.01
GenDot / IDefOpt / cpu / Forward 0.00001076287998330372 s 0.000010992019988407264 s 0.98
GenDot / JaXPipe / cpu / PreRev 0.000011555180001323609 s 0.000011150480022479314 s 1.04
GenDot / JaXPipe / cpu / PostRev 0.000010267519994613397 s 0.0000100475599992933 s 1.02
GenDot / JaXPipe / cpu / BothRev 0.000011275719991772348 s 0.000011710220014720108 s 0.96
GenDot / Jax / cpu / BothRev 0.000010862500012081 s 0.000010871300009966944 s 1.00
GenDot / HLOOpt / cpu / PreRev 0.000011150639991228671 s 0.00001220950001879828 s 0.91
GenDot / HLOOpt / cpu / PostRev 0.00001584234000347351 s 0.000013349500022741267 s 1.19
GenDot / HLOOpt / cpu / BothRev 0.000011160660001223733 s 0.000011031640015062296 s 1.01
GenDot / PartOpt / cpu / PreRev 0.000010779800006730512 s 0.000010969980039590154 s 0.98
GenDot / PartOpt / cpu / PostRev 0.000010334479995890433 s 0.000010812399959831965 s 0.96
GenDot / PartOpt / cpu / BothRev 0.000011369480007488165 s 0.000011814540011982898 s 0.96
GenDot / IPartOpt / cpu / PreRev 0.000010941820009975343 s 0.000010740520019680844 s 1.02
GenDot / IPartOpt / cpu / PostRev 0.000010436160016524809 s 0.00000975463995928294 s 1.07
GenDot / IPartOpt / cpu / BothRev 0.00001086285999917891 s 0.000011324979986966356 s 0.96
GenDot / DefOpt / cpu / PreRev 0.00001162830000566828 s 0.000011239259993089946 s 1.03
GenDot / DefOpt / cpu / PostRev 0.000011552400005712115 s 0.00001121968001825735 s 1.03
GenDot / DefOpt / cpu / BothRev 0.000011356779998550337 s 0.000011404920005588792 s 1.00
GenDot / IDefOpt / cpu / PreRev 0.000010907340001722332 s 0.0000108012400323787 s 1.01
GenDot / IDefOpt / cpu / PostRev 0.000011059800008297317 s 0.000012132959973314428 s 0.91
GenDot / IDefOpt / cpu / BothRev 0.000011521160004122069 s 0.00001086260001102346 s 1.06
GenDot / JaXPipe / cuda / Primal 0.000002527 s 0.000002015 s 1.25
GenDot / Jax / cuda / Primal 0.000002528 s 0.000002016 s 1.25
GenDot / HLOOpt / cuda / Primal 0.000002527 s 0.000001984 s 1.27
GenDot / PartOpt / cuda / Primal 0.00000256 s 0.000002016 s 1.27
GenDot / IPartOpt / cuda / Primal 0.00000256 s 0.000002016 s 1.27
GenDot / DefOpt / cuda / Primal 0.000002528 s 0.000002016 s 1.25
GenDot / IDefOpt / cuda / Primal 0.000002528 s 0.000002015 s 1.25
GenDot / JaXPipe / cuda / Forward 0.000010752 s 0.00001024 s 1.05
GenDot / Jax / cuda / Forward 0.000010688 s 0.000009983 s 1.07
GenDot / HLOOpt / cuda / Forward 0.000011809 s 0.000009856 s 1.20
GenDot / PartOpt / cuda / Forward 0.00001056 s 0.000009856 s 1.07
GenDot / IPartOpt / cuda / Forward 0.000010592 s 0.000009824 s 1.08
GenDot / DefOpt / cuda / Forward 0.000010656 s 0.00000992 s 1.07
GenDot / IDefOpt / cuda / Forward 0.000011232 s 0.000009728 s 1.15
GenDot / JaXPipe / cuda / PreRev 0.000010528 s 0.000010112 s 1.04
GenDot / JaXPipe / cuda / PostRev 0.000011647 s 0.000010016 s 1.16
GenDot / JaXPipe / cuda / BothRev 0.000010912 s 0.000010048 s 1.09
GenDot / Jax / cuda / BothRev 0.000010944 s 0.000010144 s 1.08
GenDot / HLOOpt / cuda / PreRev 0.000010688 s 0.000010016 s 1.07
GenDot / HLOOpt / cuda / PostRev 0.000010528 s 0.000010048 s 1.05
GenDot / HLOOpt / cuda / BothRev 0.000014784 s 0.000009664 s 1.53
GenDot / PartOpt / cuda / PreRev 0.000012512 s 0.000010016 s 1.25
GenDot / PartOpt / cuda / PostRev 0.000011007 s 0.00001008 s 1.09
GenDot / PartOpt / cuda / BothRev 0.000010752 s 0.000010592 s 1.02
GenDot / IPartOpt / cuda / PreRev 0.00001072 s 0.000011264 s 0.95
GenDot / IPartOpt / cuda / PostRev 0.000010912 s 0.000011648 s 0.94
GenDot / IPartOpt / cuda / BothRev 0.000010591 s 0.000011328 s 0.93
GenDot / DefOpt / cuda / PreRev 0.000010592 s 0.000011520000000000002 s 0.92
GenDot / DefOpt / cuda / PostRev 0.000010687 s 0.000010527 s 1.02
GenDot / DefOpt / cuda / BothRev 0.000010848 s 0.000011392 s 0.95
GenDot / IDefOpt / cuda / PreRev 0.000010624 s 0.000010048 s 1.06
GenDot / IDefOpt / cuda / PostRev 0.000010432 s 0.000009984 s 1.04
GenDot / IDefOpt / cuda / BothRev 0.000010368 s 0.000009792 s 1.06
GenDot / JaXPipe / tpu / Primal 9.5295e-7 s 9.30525e-7 s 1.02
GenDot / Jax / tpu / Primal 9.25925e-7 s 9.257e-7 s 1.00
GenDot / HLOOpt / tpu / Primal 0.00000165495 s 0.0000015812249999999998 s 1.05
GenDot / PartOpt / tpu / Primal 9.2605e-7 s 9.25825e-7 s 1.00
GenDot / IPartOpt / tpu / Primal 9.53275e-7 s 9.30225e-7 s 1.02
GenDot / DefOpt / tpu / Primal 0.000001507375 s 0.00000150005 s 1.00
GenDot / IDefOpt / tpu / Primal 0.00000166255 s 0.0000015821 s 1.05
GenDot / JaXPipe / tpu / Forward 0.0000030963249999999995 s 0.000003177125 s 0.97
GenDot / Jax / tpu / Forward 0.000002297025 s 0.00000232435 s 0.99
GenDot / HLOOpt / tpu / Forward 0.0000031498 s 0.00000313075 s 1.01
GenDot / PartOpt / tpu / Forward 0.00000311905 s 0.000003227975 s 0.97
GenDot / IPartOpt / tpu / Forward 0.000003151625 s 0.0000031252 s 1.01
GenDot / DefOpt / tpu / Forward 0.0000031165 s 0.000003226375 s 0.97
GenDot / IDefOpt / tpu / Forward 0.000003147975 s 0.000003131175 s 1.01
GenDot / JaXPipe / tpu / PreRev 0.0000029987500000000003 s 0.0000029877 s 1.00
GenDot / JaXPipe / tpu / PostRev 0.0000023635000000000003 s 0.000002404625 s 0.98
GenDot / JaXPipe / tpu / BothRev 0.000002983025 s 0.000002984075 s 1.00
GenDot / Jax / tpu / BothRev 0.0000023755 s 0.0000023980500000000003 s 0.99
GenDot / HLOOpt / tpu / PreRev 0.000002992325 s 0.0000029821500000000003 s 1.00
GenDot / HLOOpt / tpu / PostRev 0.0000029732500000000003 s 0.000002923425 s 1.02
GenDot / HLOOpt / tpu / BothRev 0.00000298905 s 0.0000029864249999999995 s 1.00
GenDot / PartOpt / tpu / PreRev 0.000002968325 s 0.000002925775 s 1.01
GenDot / PartOpt / tpu / PostRev 0.000002400725 s 0.0000023952 s 1.00
GenDot / PartOpt / tpu / BothRev 0.000002976 s 0.000002923075 s 1.02
GenDot / IPartOpt / tpu / PreRev 0.00000299205 s 0.0000029841 s 1.00
GenDot / IPartOpt / tpu / PostRev 0.0000023682 s 0.000002404425 s 0.98
GenDot / IPartOpt / tpu / BothRev 0.00000298875 s 0.000002975 s 1.00
GenDot / DefOpt / tpu / PreRev 0.00000296805 s 0.000002927375 s 1.01
GenDot / DefOpt / tpu / PostRev 0.000002994025 s 0.00000298225 s 1.00
GenDot / DefOpt / tpu / BothRev 0.000002978925 s 0.000002920825 s 1.02
GenDot / IDefOpt / tpu / PreRev 0.000002998875 s 0.00000298085 s 1.01
GenDot / IDefOpt / tpu / PostRev 0.000002978575 s 0.0000029257 s 1.02
GenDot / IDefOpt / tpu / BothRev 0.0000029989749999999995 s 0.0000029752250000000004 s 1.01
GenDot / JaXPipe / cpu / Primal 0.000014976 s 0.000007543960009570583 s 1.99
GenDot / Jax / cpu / Primal 0.000014724 s 0.00000684336000631447 s 2.15
GenDot / HLOOpt / cpu / Primal 0.000013766 s 0.000007792519991198787 s 1.77
GenDot / PartOpt / cpu / Primal 0.000015178 s 0.000006615960010094568 s 2.29
GenDot / IPartOpt / cpu / Primal 0.000014814 s 0.000006698160013911547 s 2.21
GenDot / DefOpt / cpu / Primal 0.000013928 s 0.000007053299959807191 s 1.97
GenDot / IDefOpt / cpu / Primal 0.00001431 s 0.00000699161998454656 s 2.05
GenDot / JaXPipe / cpu / Forward 0.000019488 s 0.000010652839973772645 s 1.83
GenDot / Jax / cpu / Forward 0.000020563000000000003 s 0.000010071080032503232 s 2.04
GenDot / HLOOpt / cpu / Forward 0.000019448 s 0.000011345020020598896 s 1.71
GenDot / PartOpt / cpu / Forward 0.000019223 s 0.000010618520036587142 s 1.81
GenDot / IPartOpt / cpu / Forward 0.000018927 s 0.000011141940003653872 s 1.70
GenDot / DefOpt / cpu / Forward 0.000019315 s 0.000011015899963240372 s 1.75
GenDot / IDefOpt / cpu / Forward 0.000019091 s 0.000010992019988407264 s 1.74
GenDot / JaXPipe / cpu / PreRev 0.000019179 s 0.000011150480022479314 s 1.72
GenDot / JaXPipe / cpu / PostRev 0.000020113 s 0.0000100475599992933 s 2.00
GenDot / JaXPipe / cpu / BothRev 0.000019041 s 0.000011710220014720108 s 1.63
GenDot / Jax / cpu / BothRev 0.000020298 s 0.000010871300009966944 s 1.87
GenDot / HLOOpt / cpu / PreRev 0.00001984 s 0.00001220950001879828 s 1.62
GenDot / HLOOpt / cpu / PostRev 0.000019152 s 0.000013349500022741267 s 1.43
GenDot / HLOOpt / cpu / BothRev 0.000019076 s 0.000011031640015062296 s 1.73
GenDot / PartOpt / cpu / PreRev 0.00001884 s 0.000010969980039590154 s 1.72
GenDot / PartOpt / cpu / PostRev 0.000019734 s 0.000010812399959831965 s 1.83
GenDot / PartOpt / cpu / BothRev 0.000019147 s 0.000011814540011982898 s 1.62
GenDot / IPartOpt / cpu / PreRev 0.0000197 s 0.000010740520019680844 s 1.83
GenDot / IPartOpt / cpu / PostRev 0.000020572 s 0.00000975463995928294 s 2.11
GenDot / IPartOpt / cpu / BothRev 0.000019072 s 0.000011324979986966356 s 1.68
GenDot / DefOpt / cpu / PreRev 0.000018818 s 0.000011239259993089946 s 1.67
GenDot / DefOpt / cpu / PostRev 0.000019226 s 0.00001121968001825735 s 1.71
GenDot / DefOpt / cpu / BothRev 0.000019523 s 0.000011404920005588792 s 1.71
GenDot / IDefOpt / cpu / PreRev 0.00001918 s 0.0000108012400323787 s 1.78
GenDot / IDefOpt / cpu / PostRev 0.0000194 s 0.000012132959973314428 s 1.60
GenDot / IDefOpt / cpu / BothRev 0.000019235 s 0.00001086260001102346 s 1.77
hlo_ffi / JaXPipe / cpu / Primal 0.00001038149999885718 s 0.00001027235996843956 s 1.01
hlo_ffi / Jax / cpu / Primal 0.000010573280007974972 s 0.00001028005999614834 s 1.03
hlo_ffi / HLOOpt / cpu / Primal 0.000009918780003772554 s 0.000010306739959560218 s 0.96
hlo_ffi / PartOpt / cpu / Primal 0.000009689539997452811 s 0.000010014420022343984 s 0.97
hlo_ffi / IPartOpt / cpu / Primal 0.000010272520003127285 s 0.000010706400044000476 s 0.96
hlo_ffi / DefOpt / cpu / Primal 0.000009719360002691248 s 0.0000098558800345927 s 0.99
hlo_ffi / IDefOpt / cpu / Primal 0.00001021986000750985 s 0.00000984211998002138 s 1.04
hlo_ffi / JaXPipe / cpu / Forward 0.000014514959993903177 s 0.00001452496000638348 s 1.00
hlo_ffi / Jax / cpu / Forward 0.000014301740009159404 s 0.000014700679985253372 s 0.97
hlo_ffi / HLOOpt / cpu / Forward 0.000014562479989308486 s 0.00001474948001487064 s 0.99
hlo_ffi / PartOpt / cpu / Forward 0.000014562499989096978 s 0.000014515399998344948 s 1.00
hlo_ffi / IPartOpt / cpu / Forward 0.00001482722000218928 s 0.000014186420003170497 s 1.05
hlo_ffi / DefOpt / cpu / Forward 0.000014444320001985032 s 0.000014972280032452544 s 0.96
hlo_ffi / IDefOpt / cpu / Forward 0.000014567160001206504 s 0.000014576040011888836 s 1.00
hlo_ffi / JaXPipe / cpu / PreRev 0.000015176319996044184 s 0.000014949839978726232 s 1.02
hlo_ffi / JaXPipe / cpu / PostRev 0.000014340599998377 s 0.000014624340046793804 s 0.98
hlo_ffi / JaXPipe / cpu / BothRev 0.000013647860000673971 s 0.000013843159995303725 s 0.99
hlo_ffi / Jax / cpu / BothRev 0.000014636260002589552 s 0.000015019619995655376 s 0.97
hlo_ffi / HLOOpt / cpu / PreRev 0.000014478439998129034 s 0.00001564962006341375 s 0.93
hlo_ffi / HLOOpt / cpu / PostRev 0.000016292180014261248 s 0.000016052320006565424 s 1.01
hlo_ffi / HLOOpt / cpu / BothRev 0.000013957319995370198 s 0.00001432220001333917 s 0.97
hlo_ffi / PartOpt / cpu / PreRev 0.000014288479994775115 s 0.00001576888003910426 s 0.91
hlo_ffi / PartOpt / cpu / PostRev 0.000013744840000526892 s 0.00001419655999598035 s 0.97
hlo_ffi / PartOpt / cpu / BothRev 0.000014100860000780812 s 0.000014534319980157308 s 0.97
hlo_ffi / IPartOpt / cpu / PreRev 0.000014443379993736017 s 0.000015157460038608406 s 0.95
hlo_ffi / IPartOpt / cpu / PostRev 0.000014146639991849953 s 0.000014356760048030991 s 0.99
hlo_ffi / IPartOpt / cpu / BothRev 0.000014187999997830048 s 0.000014438420002989003 s 0.98
hlo_ffi / DefOpt / cpu / PreRev 0.000014876160012136096 s 0.000014575639988834157 s 1.02
hlo_ffi / DefOpt / cpu / PostRev 0.000014468460003627116 s 0.000014225739969333515 s 1.02
hlo_ffi / DefOpt / cpu / BothRev 0.00001418334000845789 s 0.000014324239991765353 s 0.99
hlo_ffi / IDefOpt / cpu / PreRev 0.000014425900005790026 s 0.00001521861997389351 s 0.95
hlo_ffi / IDefOpt / cpu / PostRev 0.00001403898000035042 s 0.000014588500034733444 s 0.96
hlo_ffi / IDefOpt / cpu / BothRev 0.00001408223999760594 s 0.000014252720002332352 s 0.99
hlo_ffi / JaXPipe / cuda / Primal 0.0000023670000000000004 s 0.000001984 s 1.19
hlo_ffi / Jax / cuda / Primal 0.000002368 s 0.000001983 s 1.19
hlo_ffi / HLOOpt / cuda / Primal 0.0000023670000000000004 s 0.000001984 s 1.19
hlo_ffi / PartOpt / cuda / Primal 0.000002368 s 0.000001984 s 1.19
hlo_ffi / IPartOpt / cuda / Primal 0.0000023670000000000004 s 0.000001984 s 1.19
hlo_ffi / DefOpt / cuda / Primal 0.000002368 s 0.000001984 s 1.19
hlo_ffi / IDefOpt / cuda / Primal 0.0000023670000000000004 s 0.000001983 s 1.19
hlo_ffi / JaXPipe / cuda / Forward 0.000002463 s 0.00000208 s 1.18
hlo_ffi / Jax / cuda / Forward 0.000002463 s 0.00000208 s 1.18
hlo_ffi / HLOOpt / cuda / Forward 0.000002463 s 0.00000208 s 1.18
hlo_ffi / PartOpt / cuda / Forward 0.000002463 s 0.00000208 s 1.18
hlo_ffi / IPartOpt / cuda / Forward 0.000002463 s 0.00000208 s 1.18
hlo_ffi / DefOpt / cuda / Forward 0.000002431 s 0.00000208 s 1.17
hlo_ffi / IDefOpt / cuda / Forward 0.000002463 s 0.00000208 s 1.18
hlo_ffi / JaXPipe / cuda / PreRev 0.000002432 s 0.000002048 s 1.19
hlo_ffi / JaXPipe / cuda / PostRev 0.000002463 s 0.000002048 s 1.20
hlo_ffi / JaXPipe / cuda / BothRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / Jax / cuda / BothRev 0.000002463 s 0.00000208 s 1.18
hlo_ffi / HLOOpt / cuda / PreRev 0.000002432 s 0.000002048 s 1.19
hlo_ffi / HLOOpt / cuda / PostRev 0.000002463 s 0.000002048 s 1.20
hlo_ffi / HLOOpt / cuda / BothRev 0.000002432 s 0.000002048 s 1.19
hlo_ffi / PartOpt / cuda / PreRev 0.000002463 s 0.000002048 s 1.20
hlo_ffi / PartOpt / cuda / PostRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / PartOpt / cuda / BothRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IPartOpt / cuda / PreRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IPartOpt / cuda / PostRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / IPartOpt / cuda / BothRev 0.000002432 s 0.000002048 s 1.19
hlo_ffi / DefOpt / cuda / PreRev 0.000002464 s 0.00000208 s 1.18
hlo_ffi / DefOpt / cuda / PostRev 0.000002432 s 0.000002048 s 1.19
hlo_ffi / DefOpt / cuda / BothRev 0.000002432 s 0.000002048 s 1.19
hlo_ffi / IDefOpt / cuda / PreRev 0.000002431 s 0.000002047 s 1.19
hlo_ffi / IDefOpt / cuda / PostRev 0.000002432 s 0.000002048 s 1.19
hlo_ffi / IDefOpt / cuda / BothRev 0.000002463 s 0.000002048 s 1.20
hlo_ffi / JaXPipe / tpu / Primal 9.21125e-7 s 9.172e-7 s 1.00
hlo_ffi / Jax / tpu / Primal 9.50475e-7 s 9.499e-7 s 1.00
hlo_ffi / HLOOpt / tpu / Primal 8.953999999999999e-7 s 8.99325e-7 s 1.00
hlo_ffi / PartOpt / tpu / Primal 9.5055e-7 s 9.55225e-7 s 1.00
hlo_ffi / IPartOpt / tpu / Primal 8.994e-7 s 9.028e-7 s 1.00
hlo_ffi / DefOpt / tpu / Primal 9.52775e-7 s 9.5425e-7 s 1.00
hlo_ffi / IDefOpt / tpu / Primal 9.00975e-7 s 8.985500000000001e-7 s 1.00
hlo_ffi / JaXPipe / tpu / Forward 9.4945e-7 s 9.49375e-7 s 1.00
hlo_ffi / Jax / tpu / Forward 9.82125e-7 s 9.8175e-7 s 1.00
hlo_ffi / HLOOpt / tpu / Forward 9.74775e-7 s 9.7415e-7 s 1.00
hlo_ffi / PartOpt / tpu / Forward 9.34725e-7 s 9.3425e-7 s 1.00
hlo_ffi / IPartOpt / tpu / Forward 9.74675e-7 s 9.73925e-7 s 1.00
hlo_ffi / DefOpt / tpu / Forward 9.34425e-7 s 9.34025e-7 s 1.00
hlo_ffi / IDefOpt / tpu / Forward 9.74525e-7 s 9.746e-7 s 1.00
hlo_ffi / JaXPipe / tpu / PreRev 9.31575e-7 s 9.323e-7 s 1.00
hlo_ffi / JaXPipe / tpu / PostRev 9.6565e-7 s 9.654e-7 s 1.00
hlo_ffi / JaXPipe / tpu / BothRev 9.626e-7 s 9.63e-7 s 1.00
hlo_ffi / Jax / tpu / BothRev 9.65525e-7 s 9.653e-7 s 1.00
hlo_ffi / HLOOpt / tpu / PreRev 9.629e-7 s 9.6215e-7 s 1.00
hlo_ffi / HLOOpt / tpu / PostRev 9.65825e-7 s 9.65175e-7 s 1.00
hlo_ffi / HLOOpt / tpu / BothRev 9.629e-7 s 9.62175e-7 s 1.00
hlo_ffi / PartOpt / tpu / PreRev 9.65325e-7 s 9.648e-7 s 1.00
hlo_ffi / PartOpt / tpu / PostRev 9.62125e-7 s 9.62225e-7 s 1.00
hlo_ffi / PartOpt / tpu / BothRev 9.653e-7 s 9.654e-7 s 1.00
hlo_ffi / IPartOpt / tpu / PreRev 9.625250000000002e-7 s 9.6265e-7 s 1.00
hlo_ffi / IPartOpt / tpu / PostRev 9.6535e-7 s 9.652e-7 s 1.00
hlo_ffi / IPartOpt / tpu / BothRev 9.62625e-7 s 9.619e-7 s 1.00
hlo_ffi / DefOpt / tpu / PreRev 9.65225e-7 s 9.6525e-7 s 1.00
hlo_ffi / DefOpt / tpu / PostRev 9.62225e-7 s 9.62225e-7 s 1
hlo_ffi / DefOpt / tpu / BothRev 9.658e-7 s 9.65175e-7 s 1.00
hlo_ffi / IDefOpt / tpu / PreRev 9.62025e-7 s 9.6225e-7 s 1.00
hlo_ffi / IDefOpt / tpu / PostRev 9.655e-7 s 9.6515e-7 s 1.00
hlo_ffi / IDefOpt / tpu / BothRev 9.625250000000002e-7 s 9.62225e-7 s 1.00
hlo_ffi / JaXPipe / cpu / Primal 0.00001761 s 0.00001027235996843956 s 1.71
hlo_ffi / Jax / cpu / Primal 0.000017165 s 0.00001028005999614834 s 1.67
hlo_ffi / HLOOpt / cpu / Primal 0.000016979 s 0.000010306739959560218 s 1.65
hlo_ffi / PartOpt / cpu / Primal 0.000017579999999999998 s 0.000010014420022343984 s 1.76
hlo_ffi / IPartOpt / cpu / Primal 0.000017035000000000002 s 0.000010706400044000476 s 1.59
hlo_ffi / DefOpt / cpu / Primal 0.00001706 s 0.0000098558800345927 s 1.73
hlo_ffi / IDefOpt / cpu / Primal 0.000017358 s 0.00000984211998002138 s 1.76
hlo_ffi / JaXPipe / cpu / Forward 0.000024857 s 0.00001452496000638348 s 1.71
hlo_ffi / Jax / cpu / Forward 0.000023863 s 0.000014700679985253372 s 1.62
hlo_ffi / HLOOpt / cpu / Forward 0.000024232 s 0.00001474948001487064 s 1.64
hlo_ffi / PartOpt / cpu / Forward 0.000023826 s 0.000014515399998344948 s 1.64
hlo_ffi / IPartOpt / cpu / Forward 0.000024071 s 0.000014186420003170497 s 1.70
hlo_ffi / DefOpt / cpu / Forward 0.000024472 s 0.000014972280032452544 s 1.63
hlo_ffi / IDefOpt / cpu / Forward 0.000024324 s 0.000014576040011888836 s 1.67
hlo_ffi / JaXPipe / cpu / PreRev 0.000024632 s 0.000014949839978726232 s 1.65
hlo_ffi / JaXPipe / cpu / PostRev 0.000023728 s 0.000014624340046793804 s 1.62
hlo_ffi / JaXPipe / cpu / BothRev 0.000023156 s 0.000013843159995303725 s 1.67
hlo_ffi / Jax / cpu / BothRev 0.000023863 s 0.000015019619995655376 s 1.59
hlo_ffi / HLOOpt / cpu / PreRev 0.000023501 s 0.00001564962006341375 s 1.50
hlo_ffi / HLOOpt / cpu / PostRev 0.000023868 s 0.000016052320006565424 s 1.49
hlo_ffi / HLOOpt / cpu / BothRev 0.000024206 s 0.00001432220001333917 s 1.69
hlo_ffi / PartOpt / cpu / PreRev 0.000024116 s 0.00001576888003910426 s 1.53
hlo_ffi / PartOpt / cpu / PostRev 0.000023601 s 0.00001419655999598035 s 1.66
hlo_ffi / PartOpt / cpu / BothRev 0.000023454 s 0.000014534319980157308 s 1.61
hlo_ffi / IPartOpt / cpu / PreRev 0.000023574 s 0.000015157460038608406 s 1.56
hlo_ffi / IPartOpt / cpu / PostRev 0.000023908 s 0.000014356760048030991 s 1.67
hlo_ffi / IPartOpt / cpu / BothRev 0.000023736 s 0.000014438420002989003 s 1.64
hlo_ffi / DefOpt / cpu / PreRev 0.000023722 s 0.000014575639988834157 s 1.63
hlo_ffi / DefOpt / cpu / PostRev 0.000023855 s 0.000014225739969333515 s 1.68
hlo_ffi / DefOpt / cpu / BothRev 0.000023973 s 0.000014324239991765353 s 1.67
hlo_ffi / IDefOpt / cpu / PreRev 0.000023793 s 0.00001521861997389351 s 1.56
hlo_ffi / IDefOpt / cpu / PostRev 0.000023866 s 0.000014588500034733444 s 1.64
hlo_ffi / IDefOpt / cpu / BothRev 0.000023945 s 0.000014252720002332352 s 1.68
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Primal 0.0009538762000147 s 0.0009321387999079 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Primal 0.0009605385999293 s 0.0009321677998741 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Primal 0.0010455776000071 s 0.0009918829998241 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Primal 0.0010059096000077 s 0.0009257460001208 s 1.09
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Primal 0.0009544355999878 s 0.0009386104000441 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Primal 0.0010204394000311 s 0.0010159395999835 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Primal 0.0010536388000218 s 0.0010079241998937 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Forward 0.002493666600003 s 0.0022614704000261 s 1.10
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Forward 0.0024664087999781 s 0.002405369400094 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Forward 0.0024128262000431 s 0.0022799684000347 s 1.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Forward 0.0023066749999998 s 0.0022509202000946 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Forward 0.0023597969999855 s 0.0022421316000873 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Forward 0.0024723510000285 s 0.0022940865999771 s 1.08
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Forward 0.0025821447999987 s 0.0022375612001269 s 1.15
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PreRev 0.005694930799973 s 0.0056978842000717 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PostRev 0.005460876200027 s 0.0063587201999325 s 0.86
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / BothRev 0.0055300228000078 s 0.0060367627999767 s 0.92
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / BothRev 0.006244140400031 s 0.0063693950001834 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PreRev 0.0059795702000201 s 0.006341317999977 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PostRev 0.0064720224000211 s 0.0053775515999404 s 1.20
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.0064234065999926 s 0.0055548254001223 s 1.16
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PreRev 0.0036307459999989 s 0.0055048000000169 s 0.66
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PostRev 0.0065919548000238 s 0.0061269304000234 s 1.08
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / BothRev 0.0035469312000032 s 0.0056070170000566 s 0.63
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PreRev 0.0061102079999955 s 0.006099837199963 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PostRev 0.0039934354000479 s 0.0040929354000581 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / BothRev 0.0052267926000013 s 0.005288503400061 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PreRev 0.0037731398000232 s 0.0034711325999523 s 1.09
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PostRev 0.0064068107999901 s 0.0060114416000942 s 1.07
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / BothRev 0.0062970649999897 s 0.0035173248000319 s 1.79
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PreRev 0.005883673199969 s 0.0034985649999725 s 1.68
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PostRev 0.0035706702000197 s 0.0056002957999226 s 0.64
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.0062693802000239 s 0.0033613042000069 s 1.87
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / Primal 0.0002949419999999 s 0.000275552 s 1.07
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / Primal 0.000294558 s 0.00027424 s 1.07
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / Primal 0.000300766 s 0.000290208 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / Primal 0.000294847 s 0.000275392 s 1.07
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / Primal 0.000294814 s 0.000276576 s 1.07
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / Primal 0.000302111 s 0.000291777 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / Primal 0.000300862 s 0.000291072 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / Forward 0.0005823009999999 s 0.0005613759999999 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / Forward 0.000566941 s 0.000542817 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / Forward 0.000583005 s 0.000560545 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / Forward 0.000582333 s 0.000561408 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / Forward 0.00058262 s 0.000560993 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / Forward 0.00058342 s 0.000561793 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / Forward 0.000582685 s 0.000561408 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / PreRev 0.001053882 s 0.0010280979999999 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / PostRev 0.0010118339999999 s 0.000987105 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / BothRev 0.001048474 s 0.001029025 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / BothRev 0.001006523 s 0.000989153 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / PreRev 0.001034362 s 0.001013761 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / PostRev 0.001059035 s 0.001039041 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / BothRev 0.001033498 s 0.001013344 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / PreRev 0.001048602 s 0.001028897 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / PostRev 0.00099785 s 0.000978369 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / BothRev 0.001048987 s 0.001029601 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / PreRev 0.001047962 s 0.0010275529999999 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / PostRev 0.000997211 s 0.0009758099999999 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / BothRev 0.001047354 s 0.001027169 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / PreRev 0.001048314 s 0.001024991 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / PostRev 0.000981434 s 0.000964161 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / BothRev 0.001049434 s 0.001026305 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / PreRev 0.00104857 s 0.001024385 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / PostRev 0.001050842 s 0.001024928 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / BothRev 0.001050874 s 0.001023745 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / Primal 0.00012943525 s 0.0001232635 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / Primal 0.000123814 s 0.0001233635 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / Primal 0.00015887625 s 0.00015185925 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / Primal 0.00013091 s 0.0001300629999999 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / Primal 0.0001373765 s 0.000130404 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / Primal 0.0001452045 s 0.00014423625 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / Primal 0.0001567635 s 0.0001503662499999 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / Forward 0.000213348 s 0.0002128715 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / Forward 0.0002625387499999 s 0.000259804 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / Forward 0.000220616 s 0.0002129352499999 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / Forward 0.0002138972499999 s 0.0002101505 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / Forward 0.00021633825 s 0.00021305025 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / Forward 0.00021732475 s 0.0002101205 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / Forward 0.00021601025 s 0.0002127614999999 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / PreRev 0.0003579775 s 0.00035646575 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / PostRev 0.00025648875 s 0.0002569629999999 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / BothRev 0.0003574 s 0.0003556017499999 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / BothRev 0.0002571195 s 0.000257003 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / PreRev 0.00035768625 s 0.0003559415 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / PostRev 0.00029117175 s 0.00029076975 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / BothRev 0.0003576624999999 s 0.00035559025 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / PreRev 0.00035521225 s 0.00035583025 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / PostRev 0.00027432975 s 0.00027396675 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / BothRev 0.0003555705 s 0.00035568425 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / PreRev 0.0003572825 s 0.0003553484999999 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / PostRev 0.00027209625 s 0.00027274175 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / BothRev 0.000357457 s 0.0003554765 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / PreRev 0.0003576125 s 0.00035775425 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / PostRev 0.0002850734999999 s 0.00028377575 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / BothRev 0.00035742825 s 0.0003579029999999 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / PreRev 0.000360201 s 0.00035837225 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / PostRev 0.000301249 s 0.0003013565 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / BothRev 0.0003602795 s 0.0003580575 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Primal 0.001754068 s 0.0009321387999079 s 1.88
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Primal 0.001765103 s 0.0009321677998741 s 1.89
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Primal 0.002420108 s 0.0009918829998241 s 2.44
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Primal 0.002202452 s 0.0009257460001208 s 2.38
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Primal 0.001844952 s 0.0009386104000441 s 1.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Primal 0.001440496 s 0.0010159395999835 s 1.42
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Primal 0.001887511 s 0.0010079241998937 s 1.87
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Forward 0.0047922719999999 s 0.0022614704000261 s 2.12
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Forward 0.00490088 s 0.002405369400094 s 2.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Forward 0.0048674289999999 s 0.0022799684000347 s 2.13
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Forward 0.004681857 s 0.0022509202000946 s 2.08
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Forward 0.004818217 s 0.0022421316000873 s 2.15
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Forward 0.005358894 s 0.0022940865999771 s 2.34
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Forward 0.005434115 s 0.0022375612001269 s 2.43
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PreRev 0.007809917 s 0.0056978842000717 s 1.37
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PostRev 0.009613276 s 0.0063587201999325 s 1.51
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / BothRev 0.007504271 s 0.0060367627999767 s 1.24
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / BothRev 0.008013709 s 0.0063693950001834 s 1.26
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PreRev 0.007624387 s 0.006341317999977 s 1.20
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PostRev 0.0070181079999999 s 0.0053775515999404 s 1.31
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.008287688 s 0.0055548254001223 s 1.49
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PreRev 0.007096286 s 0.0055048000000169 s 1.29
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PostRev 0.008054555 s 0.0061269304000234 s 1.31
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / BothRev 0.007168387 s 0.0056070170000566 s 1.28
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PreRev 0.00782654 s 0.006099837199963 s 1.28
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PostRev 0.007973227 s 0.0040929354000581 s 1.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / BothRev 0.007602781 s 0.005288503400061 s 1.44
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PreRev 0.007637661 s 0.0034711325999523 s 2.20
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PostRev 0.0066090359999999 s 0.0060114416000942 s 1.10
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / BothRev 0.007865277 s 0.0035173248000319 s 2.24
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PreRev 0.008334201 s 0.0034985649999725 s 2.38
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PostRev 0.007737976 s 0.0056002957999226 s 1.38
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.00805485 s 0.0033613042000069 s 2.40
scatter_sum / JaXPipe / cpu / Primal 0.000008232799996221728 s 0.00000753280001845269 s 1.09
scatter_sum / Jax / cpu / Primal 0.0000074311600019427716 s 0.000008337480048794532 s 0.89
scatter_sum / HLOOpt / cpu / Primal 0.000007944459998725506 s 0.000007767980041535339 s 1.02
scatter_sum / PartOpt / cpu / Primal 0.000007805500010817924 s 0.00000764754005103896 s 1.02
scatter_sum / IPartOpt / cpu / Primal 0.000007823479993476212 s 0.000008083640050244867 s 0.97
scatter_sum / DefOpt / cpu / Primal 0.000007532980002906698 s 0.00000791293996371678 s 0.95
scatter_sum / IDefOpt / cpu / Primal 0.000007370640012140939 s 0.000007435879997501615 s 0.99
scatter_sum / JaXPipe / cpu / Forward 0.000011610780004502886 s 0.000012307519973546732 s 0.94
scatter_sum / Jax / cpu / Forward 0.000011827940004423 s 0.00001249038001333247 s 0.95
scatter_sum / HLOOpt / cpu / Forward 0.00001236350000681341 s 0.000012609119976332294 s 0.98
scatter_sum / PartOpt / cpu / Forward 0.000011492219994124752 s 0.00001196134005112981 s 0.96
scatter_sum / IPartOpt / cpu / Forward 0.000012607880003088211 s 0.000012671400008912317 s 0.99
scatter_sum / DefOpt / cpu / Forward 0.00001185186001293914 s 0.000012489519986047524 s 0.95
scatter_sum / IDefOpt / cpu / Forward 0.000011751879997063952 s 0.00001238691998878494 s 0.95
scatter_sum / JaXPipe / cpu / PreRev 0.000012054400001488827 s 0.000011618060007094757 s 1.04
scatter_sum / JaXPipe / cpu / PostRev 0.000011956700000155251 s 0.000012251960006324224 s 0.98
scatter_sum / JaXPipe / cpu / BothRev 0.000012448499999209162 s 0.00001236012001754716 s 1.01
scatter_sum / Jax / cpu / BothRev 0.00001176456000393955 s 0.00001187192001452786 s 0.99
scatter_sum / HLOOpt / cpu / PreRev 0.000011926420008876447 s 0.000012426240036802485 s 0.96
scatter_sum / HLOOpt / cpu / PostRev 0.000014032039998710387 s 0.000014015159995324212 s 1.00
scatter_sum / HLOOpt / cpu / BothRev 0.00001215489999140118 s 0.000011412799985919263 s 1.07
scatter_sum / PartOpt / cpu / PreRev 0.000011534219997884066 s 0.000011961099971813384 s 0.96
scatter_sum / PartOpt / cpu / PostRev 0.000011970179996296792 s 0.000011408780037527322 s 1.05
scatter_sum / PartOpt / cpu / BothRev 0.000012949399999797608 s 0.000012410079989422227 s 1.04
scatter_sum / IPartOpt / cpu / PreRev 0.000011889779991633986 s 0.000012210839995532295 s 0.97
scatter_sum / IPartOpt / cpu / PostRev 0.000011943519998567354 s 0.000011338740032442729 s 1.05
scatter_sum / IPartOpt / cpu / BothRev 0.00001298193999900832 s 0.000011471719999462948 s 1.13
scatter_sum / DefOpt / cpu / PreRev 0.000012024759998894296 s 0.000011868340016008003 s 1.01
scatter_sum / DefOpt / cpu / PostRev 0.000012149619990395876 s 0.000011632699988695096 s 1.04
scatter_sum / DefOpt / cpu / BothRev 0.000012437460004548484 s 0.000011750099965865956 s 1.06
scatter_sum / IDefOpt / cpu / PreRev 0.000011138120000850905 s 0.00001182487998448778 s 0.94
scatter_sum / IDefOpt / cpu / PostRev 0.00001213508000546426 s 0.000011796299995694426 s 1.03
scatter_sum / IDefOpt / cpu / BothRev 0.000012289420001252436 s 0.00001179338001747965 s 1.04
scatter_sum / JaXPipe / cuda / Primal 0.000010464 s 0.000010112 s 1.03
scatter_sum / Jax / cuda / Primal 0.000010784 s 0.000009824 s 1.10
scatter_sum / HLOOpt / cuda / Primal 0.000010432 s 0.000010112 s 1.03
scatter_sum / PartOpt / cuda / Primal 0.000010368 s 0.00001024 s 1.01
scatter_sum / IPartOpt / cuda / Primal 0.000010528 s 0.000010304 s 1.02
scatter_sum / DefOpt / cuda / Primal 0.000010751 s 0.000009791 s 1.10
scatter_sum / IDefOpt / cuda / Primal 0.000010752 s 0.000010239 s 1.05
scatter_sum / JaXPipe / cuda / Forward 0.000017215 s 0.000016896000000000002 s 1.02
scatter_sum / Jax / cuda / Forward 0.000017056 s 0.000017088 s 1.00
scatter_sum / HLOOpt / cuda / Forward 0.000017344 s 0.000017312 s 1.00
scatter_sum / PartOpt / cuda / Forward 0.000016927999999999998 s 0.000017056 s 0.99
scatter_sum / IPartOpt / cuda / Forward 0.000017408 s 0.000017152 s 1.01
scatter_sum / DefOpt / cuda / Forward 0.00001728 s 0.000016929 s 1.02
scatter_sum / IDefOpt / cuda / Forward 0.000017216 s 0.00002448 s 0.70
scatter_sum / JaXPipe / cuda / PreRev 0.000016896000000000002 s 0.000017503999999999997 s 0.97
scatter_sum / JaXPipe / cuda / PostRev 0.000016896000000000002 s 0.000016737 s 1.01
scatter_sum / JaXPipe / cuda / BothRev 0.000016927999999999998 s 0.000017152 s 0.99
scatter_sum / Jax / cuda / BothRev 0.00001712 s 0.000018753 s 0.91
scatter_sum / HLOOpt / cuda / PreRev 0.000017024 s 0.000017344 s 0.98
scatter_sum / HLOOpt / cuda / PostRev 0.000018432 s 0.000017312 s 1.06
scatter_sum / HLOOpt / cuda / BothRev 0.0000168 s 0.000017760000000000003 s 0.95
scatter_sum / PartOpt / cuda / PreRev 0.000019328 s 0.000018048 s 1.07
scatter_sum / PartOpt / cuda / PostRev 0.000016448000000000002 s 0.000017760000000000003 s 0.93
scatter_sum / PartOpt / cuda / BothRev 0.000018912 s 0.000018016 s 1.05
scatter_sum / IPartOpt / cuda / PreRev 0.000017023 s 0.000017888999999999998 s 0.95
scatter_sum / IPartOpt / cuda / PostRev 0.000016864 s 0.000016670999999999997 s 1.01
scatter_sum / IPartOpt / cuda / BothRev 0.000017344 s 0.000017888999999999998 s 0.97
scatter_sum / DefOpt / cuda / PreRev 0.000017503999999999997 s 0.000017152 s 1.02
scatter_sum / DefOpt / cuda / PostRev 0.000017152 s 0.000017152 s 1
scatter_sum / DefOpt / cuda / BothRev 0.000017888000000000002 s 0.000017216 s 1.04
scatter_sum / IDefOpt / cuda / PreRev 0.000026463 s 0.00001696 s 1.56
scatter_sum / IDefOpt / cuda / PostRev 0.000017344 s 0.000017247999999999998 s 1.01
scatter_sum / IDefOpt / cuda / BothRev 0.000017919999999999998 s 0.000017345 s 1.03
scatter_sum / JaXPipe / tpu / Primal 0.00000135045 s 0.000001343275 s 1.01
scatter_sum / Jax / tpu / Primal 0.0000013430500000000002 s 0.000001404425 s 0.96
scatter_sum / HLOOpt / tpu / Primal 0.000001350875 s 0.00000134315 s 1.01
scatter_sum / PartOpt / tpu / Primal 0.000001343175 s 0.000001404425 s 0.96
scatter_sum / IPartOpt / tpu / Primal 0.000001350575 s 0.0000013434250000000002 s 1.01
scatter_sum / DefOpt / tpu / Primal 0.0000013433999999999995 s 0.00000140475 s 0.96
scatter_sum / IDefOpt / tpu / Primal 0.0000013510500000000002 s 0.0000013432 s 1.01
scatter_sum / JaXPipe / tpu / Forward 0.00000269355 s 0.000002704675 s 1.00
scatter_sum / Jax / tpu / Forward 0.000002731575 s 0.00000271835 s 1.00
scatter_sum / HLOOpt / tpu / Forward 0.000002693525 s 0.0000027015 s 1.00
scatter_sum / PartOpt / tpu / Forward 0.000002700625 s 0.00000268595 s 1.01
scatter_sum / IPartOpt / tpu / Forward 0.000002685125 s 0.0000027071 s 0.99
scatter_sum / DefOpt / tpu / Forward 0.0000027005 s 0.0000026979 s 1.00
scatter_sum / IDefOpt / tpu / Forward 0.000002684475 s 0.000002715275 s 0.99
scatter_sum / JaXPipe / tpu / PreRev 0.000002690125 s 0.0000026796750000000003 s 1.00
scatter_sum / JaXPipe / tpu / PostRev 0.00000269015 s 0.000002681375 s 1.00
scatter_sum / JaXPipe / tpu / BothRev 0.00000270845 s 0.0000026992500000000005 s 1.00
scatter_sum / Jax / tpu / BothRev 0.00000274445 s 0.000002735225 s 1.00
scatter_sum / HLOOpt / tpu / PreRev 0.00000270715 s 0.000002694525 s 1.00
scatter_sum / HLOOpt / tpu / PostRev 0.00000274255 s 0.000002740225 s 1.00
scatter_sum / HLOOpt / tpu / BothRev 0.00000271305 s 0.0000026929750000000003 s 1.01
scatter_sum / PartOpt / tpu / PreRev 0.0000027437 s 0.000002751425 s 1.00
scatter_sum / PartOpt / tpu / PostRev 0.000002712225 s 0.0000026935 s 1.01
scatter_sum / PartOpt / tpu / BothRev 0.000002736675 s 0.0000027455 s 1.00
scatter_sum / IPartOpt / tpu / PreRev 0.00000270745 s 0.0000026997500000000003 s 1.00
scatter_sum / IPartOpt / tpu / PostRev 0.000002735975 s 0.00000273565 s 1.00
scatter_sum / IPartOpt / tpu / BothRev 0.000002713825 s 0.000002711575 s 1.00
scatter_sum / DefOpt / tpu / PreRev 0.00000275165 s 0.00000274295 s 1.00
scatter_sum / DefOpt / tpu / PostRev 0.0000027049 s 0.000002690275 s 1.01
scatter_sum / DefOpt / tpu / BothRev 0.00000273595 s 0.0000027378750000000004 s 1.00
scatter_sum / IDefOpt / tpu / PreRev 0.00000271225 s 0.0000026918750000000004 s 1.01
scatter_sum / IDefOpt / tpu / PostRev 0.00000274175 s 0.00000273465 s 1.00
scatter_sum / IDefOpt / tpu / BothRev 0.0000027071250000000004 s 0.000002694975 s 1.00
scatter_sum / JaXPipe / cpu / Primal 0.000015614 s 0.00000753280001845269 s 2.07
scatter_sum / Jax / cpu / Primal 0.000015807000000000003 s 0.000008337480048794532 s 1.90
scatter_sum / HLOOpt / cpu / Primal 0.000015988 s 0.000007767980041535339 s 2.06
scatter_sum / PartOpt / cpu / Primal 0.000015683 s 0.00000764754005103896 s 2.05
scatter_sum / IPartOpt / cpu / Primal 0.000015712 s 0.000008083640050244867 s 1.94
scatter_sum / DefOpt / cpu / Primal 0.000015895 s 0.00000791293996371678 s 2.01
scatter_sum / IDefOpt / cpu / Primal 0.000016088000000000002 s 0.000007435879997501615 s 2.16
scatter_sum / JaXPipe / cpu / Forward 0.000022516000000000003 s 0.000012307519973546732 s 1.83
scatter_sum / Jax / cpu / Forward 0.000022248 s 0.00001249038001333247 s 1.78
scatter_sum / HLOOpt / cpu / Forward 0.000022615 s 0.000012609119976332294 s 1.79
scatter_sum / PartOpt / cpu / Forward 0.000022931 s 0.00001196134005112981 s 1.92
scatter_sum / IPartOpt / cpu / Forward 0.000021954 s 0.000012671400008912317 s 1.73
scatter_sum / DefOpt / cpu / Forward 0.000022358 s 0.000012489519986047524 s 1.79
scatter_sum / IDefOpt / cpu / Forward 0.00002268 s 0.00001238691998878494 s 1.83
scatter_sum / JaXPipe / cpu / PreRev 0.000023416 s 0.000011618060007094757 s 2.02
scatter_sum / JaXPipe / cpu / PostRev 0.000022020000000000003 s 0.000012251960006324224 s 1.80
scatter_sum / JaXPipe / cpu / BothRev 0.000022531 s 0.00001236012001754716 s 1.82
scatter_sum / Jax / cpu / BothRev 0.000022093 s 0.00001187192001452786 s 1.86
scatter_sum / HLOOpt / cpu / PreRev 0.000022214 s 0.000012426240036802485 s 1.79
scatter_sum / HLOOpt / cpu / PostRev 0.0000228 s 0.000014015159995324212 s 1.63
scatter_sum / HLOOpt / cpu / BothRev 0.000022513 s 0.000011412799985919263 s 1.97
scatter_sum / PartOpt / cpu / PreRev 0.000022711 s 0.000011961099971813384 s 1.90
scatter_sum / PartOpt / cpu / PostRev 0.000022609 s 0.000011408780037527322 s 1.98
scatter_sum / PartOpt / cpu / BothRev 0.000022338 s 0.000012410079989422227 s 1.80
scatter_sum / IPartOpt / cpu / PreRev 0.000021862 s 0.000012210839995532295 s 1.79
scatter_sum / IPartOpt / cpu / PostRev 0.000022485 s 0.000011338740032442729 s 1.98
scatter_sum / IPartOpt / cpu / BothRev 0.000022412 s 0.000011471719999462948 s 1.95
scatter_sum / DefOpt / cpu / PreRev 0.000022218000000000003 s 0.000011868340016008003 s 1.87
scatter_sum / DefOpt / cpu / PostRev 0.000021823 s 0.000011632699988695096 s 1.88
scatter_sum / DefOpt / cpu / BothRev 0.00002244 s 0.000011750099965865956 s 1.91
scatter_sum / IDefOpt / cpu / PreRev 0.000022333 s 0.00001182487998448778 s 1.89
scatter_sum / IDefOpt / cpu / PostRev 0.000022701 s 0.000011796299995694426 s 1.92
scatter_sum / IDefOpt / cpu / BothRev 0.000022338 s 0.00001179338001747965 s 1.89
slicing / JaXPipe / cpu / Primal 0.000006396680003035727 s 0.000006597080018764246 s 0.97
slicing / Jax / cpu / Primal 0.0000063354199983223226 s 0.000006177960012792028 s 1.03
slicing / HLOOpt / cpu / Primal 0.000006364179996580788 s 0.000006341799980873475 s 1.00
slicing / PartOpt / cpu / Primal 0.000005998199999339704 s 0.000006137679984021815 s 0.98
slicing / IPartOpt / cpu / Primal 0.000006855900005575677 s 0.000006653579976045876 s 1.03
slicing / DefOpt / cpu / Primal 0.000006347959999857266 s 0.000006134799996289075 s 1.03
slicing / IDefOpt / cpu / Primal 0.000006141400001524744 s 0.000006215719995452673 s 0.99
slicing / JaXPipe / cpu / Forward 0.000009510540000974289 s 0.00000958000001446635 s 0.99
slicing / Jax / cpu / Forward 0.000009559240004364255 s 0.00000932574000216846 s 1.03
slicing / HLOOpt / cpu / Forward 0.00001029594000101497 s 0.000009797440025067772 s 1.05
slicing / PartOpt / cpu / Forward 0.000009069839984476858 s 0.00000952871996560134 s 0.95
slicing / IPartOpt / cpu / Forward 0.00001072107999561922 s 0.00001003165997644828 s 1.07
slicing / DefOpt / cpu / Forward 0.000009490739996635966 s 0.000009296859998357831 s 1.02
slicing / IDefOpt / cpu / Forward 0.0000100604599970211 s 0.000009523819971946069 s 1.06
slicing / JaXPipe / cpu / PreRev 0.00000993513999901552 s 0.000010216460013907636 s 0.97
slicing / JaXPipe / cpu / PostRev 0.000010720899997522792 s 0.000009857619970716768 s 1.09
slicing / JaXPipe / cpu / BothRev 0.000010598640001262538 s 0.000010476039997229235 s 1.01
slicing / Jax / cpu / BothRev 0.000010536199995385687 s 0.000009995319987865516 s 1.05
slicing / HLOOpt / cpu / PreRev 0.000010471299999608163 s 0.000010662620015864376 s 0.98
slicing / HLOOpt / cpu / PostRev 0.00001180471999987276 s 0.0000123771199832845 s 0.95
slicing / HLOOpt / cpu / BothRev 0.00001043917999822952 s 0.0000095578600121371 s 1.09
slicing / PartOpt / cpu / PreRev 0.000010487279994322308 s 0.000009616839970476575 s 1.09
slicing / PartOpt / cpu / PostRev 0.000010566960002051928 s 0.0000105017599707935 s 1.01
slicing / PartOpt / cpu / BothRev 0.00001069566000069244 s 0.000010213999994448386 s 1.05
slicing / IPartOpt / cpu / PreRev 0.00001030391999620406 s 0.000010010960013460136 s 1.03
slicing / IPartOpt / cpu / PostRev 0.000010726800005613769 s 0.000010388719992988626 s 1.03
slicing / IPartOpt / cpu / BothRev 0.00001021751998905529 s 0.000009856840006250422 s 1.04
slicing / DefOpt / cpu / PreRev 0.00001040760000478258 s 0.000009915399969031567 s 1.05
slicing / DefOpt / cpu / PostRev 0.000010074699989672809 s 0.000010333379996154693 s 0.97
slicing / DefOpt / cpu / BothRev 0.00001030389998732062 s 0.000010391340001660863 s 0.99
slicing / IDefOpt / cpu / PreRev 0.000010039699998287689 s 0.00000994755997453467 s 1.01
slicing / IDefOpt / cpu / PostRev 0.000010290640013863594 s 0.00000992942003904318 s 1.04
slicing / IDefOpt / cpu / BothRev 0.000010400779999599762 s 0.000009599200056982228 s 1.08
slicing / JaXPipe / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / Jax / cuda / Primal 0.000002304 s 0.000001887 s 1.22
slicing / HLOOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / PartOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / IPartOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / DefOpt / cuda / Primal 0.000002304 s 0.000001887 s 1.22
slicing / IDefOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / JaXPipe / cuda / Forward 0.000010463 s 0.000009856 s 1.06
slicing / Jax / cuda / Forward 0.000010144 s 0.000010175 s 1.00
slicing / HLOOpt / cuda / Forward 0.0000104 s 0.000010048 s 1.04
slicing / PartOpt / cuda / Forward 0.000010272 s 0.000010176 s 1.01
slicing / IPartOpt / cuda / Forward 0.000010208 s 0.000009856 s 1.04
slicing / DefOpt / cuda / Forward 0.000010527 s 0.00000992 s 1.06
slicing / IDefOpt / cuda / Forward 0.000010272 s 0.000010145 s 1.01
slicing / JaXPipe / cuda / PreRev 0.000010368 s 0.000009856 s 1.05
slicing / JaXPipe / cuda / PostRev 0.00001024 s 0.000010431 s 0.98
slicing / JaXPipe / cuda / BothRev 0.000010304 s 0.000010112 s 1.02
slicing / Jax / cuda / BothRev 0.000010816 s 0.00001024 s 1.06
slicing / HLOOpt / cuda / PreRev 0.000012992 s 0.000009888 s 1.31
slicing / HLOOpt / cuda / PostRev 0.000009984 s 0.000009664 s 1.03
slicing / HLOOpt / cuda / BothRev 0.000010432 s 0.000009952 s 1.05
slicing / PartOpt / cuda / PreRev 0.000010144 s 0.00001008 s 1.01
slicing / PartOpt / cuda / PostRev 0.000010432 s 0.000009792 s 1.07
slicing / PartOpt / cuda / BothRev 0.00001024 s 0.000009824 s 1.04
slicing / IPartOpt / cuda / PreRev 0.000010464 s 0.00000976 s 1.07
slicing / IPartOpt / cuda / PostRev 0.000012831 s 0.000009824 s 1.31
slicing / IPartOpt / cuda / BothRev 0.000010497 s 0.000009825 s 1.07
slicing / DefOpt / cuda / PreRev 0.000010271 s 0.000009856 s 1.04
slicing / DefOpt / cuda / PostRev 0.000010272 s 0.000010048 s 1.02
slicing / DefOpt / cuda / BothRev 0.000010496 s 0.000009791 s 1.07
slicing / IDefOpt / cuda / PreRev 0.000010239 s 0.00001008 s 1.02
slicing / IDefOpt / cuda / PostRev 0.000010368 s 0.000010111 s 1.03
slicing / IDefOpt / cuda / BothRev 0.000010624 s 0.000010048 s 1.06
slicing / JaXPipe / tpu / Primal 9.972e-7 s 9.744e-7 s 1.02
slicing / Jax / tpu / Primal 9.624e-7 s 9.7205e-7 s 0.99
slicing / HLOOpt / tpu / Primal 0.00000100655 s 9.65325e-7 s 1.04
slicing / PartOpt / tpu / Primal 9.6515e-7 s 9.68125e-7 s 1.00
slicing / IPartOpt / tpu / Primal 9.95875e-7 s 9.672999999999998e-7 s 1.03
slicing / DefOpt / tpu / Primal 9.60625e-7 s 9.6845e-7 s 0.99
slicing / IDefOpt / tpu / Primal 0.000001006025 s 9.64425e-7 s 1.04
slicing / JaXPipe / tpu / Forward 0.0000014087250000000002 s 0.000001409775 s 1.00
slicing / Jax / tpu / Forward 0.00000145125 s 0.000001425125 s 1.02
slicing / HLOOpt / tpu / Forward 0.000001513675 s 0.0000015266 s 0.99
slicing / PartOpt / tpu / Forward 0.0000014670250000000002 s 0.000001438725 s 1.02
slicing / IPartOpt / tpu / Forward 0.00000151665 s 0.0000015182 s 1.00
slicing / DefOpt / tpu / Forward 0.000001475975 s 0.000001438125 s 1.03
slicing / IDefOpt / tpu / Forward 0.0000015211 s 0.00000151785 s 1.00
slicing / JaXPipe / tpu / PreRev 0.000002474425 s 0.000002385175 s 1.04
slicing / JaXPipe / tpu / PostRev 0.0000025074750000000005 s 0.0000025197 s 1.00
slicing / JaXPipe / tpu / BothRev 0.000002511425 s 0.0000024034500000000003 s 1.04
slicing / Jax / tpu / BothRev 0.000002528575 s 0.000002536525 s 1.00
slicing / HLOOpt / tpu / PreRev 0.000002494625 s 0.000002401 s 1.04
slicing / HLOOpt / tpu / PostRev 0.00000252865 s 0.0000025440750000000003 s 0.99
slicing / HLOOpt / tpu / BothRev 0.0000024985 s 0.00000239565 s 1.04
slicing / PartOpt / tpu / PreRev 0.0000025286 s 0.000002539375 s 1.00
slicing / PartOpt / tpu / PostRev 0.0000025033250000000003 s 0.0000023967250000000003 s 1.04
slicing / PartOpt / tpu / BothRev 0.00000253825 s 0.0000025396 s 1.00
slicing / IPartOpt / tpu / PreRev 0.0000025087 s 0.000002407025 s 1.04
slicing / IPartOpt / tpu / PostRev 0.00000253395 s 0.000002555175 s 0.99
slicing / IPartOpt / tpu / BothRev 0.000002503625 s 0.000002390125 s 1.05
slicing / DefOpt / tpu / PreRev 0.0000025236750000000004 s 0.000002549 s 0.99
slicing / DefOpt / tpu / PostRev 0.000002504125 s 0.00000240395 s 1.04
slicing / DefOpt / tpu / BothRev 0.00000253935 s 0.0000025439 s 1.00
slicing / IDefOpt / tpu / PreRev 0.0000025125 s 0.00000240165 s 1.05
slicing / IDefOpt / tpu / PostRev 0.0000025247500000000003 s 0.0000025458750000000004 s 0.99
slicing / IDefOpt / tpu / BothRev 0.000002507825 s 0.0000024013250000000003 s 1.04
slicing / JaXPipe / cpu / Primal 0.000012722 s 0.000006597080018764246 s 1.93
slicing / Jax / cpu / Primal 0.000012518 s 0.000006177960012792028 s 2.03
slicing / HLOOpt / cpu / Primal 0.000012368 s 0.000006341799980873475 s 1.95
slicing / PartOpt / cpu / Primal 0.000012535 s 0.000006137679984021815 s 2.04
slicing / IPartOpt / cpu / Primal 0.00001246 s 0.000006653579976045876 s 1.87
slicing / DefOpt / cpu / Primal 0.000012534 s 0.000006134799996289075 s 2.04
slicing / IDefOpt / cpu / Primal 0.000012539 s 0.000006215719995452673 s 2.02
slicing / JaXPipe / cpu / Forward 0.000016709000000000002 s 0.00000958000001446635 s 1.74
slicing / Jax / cpu / Forward 0.000016643000000000003 s 0.00000932574000216846 s 1.78
slicing / HLOOpt / cpu / Forward 0.000017028 s 0.000009797440025067772 s 1.74
slicing / PartOpt / cpu / Forward 0.000016840999999999997 s 0.00000952871996560134 s 1.77
slicing / IPartOpt / cpu / Forward 0.000016609 s 0.00001003165997644828 s 1.66
slicing / DefOpt / cpu / Forward 0.000016629000000000003 s 0.000009296859998357831 s 1.79
slicing / IDefOpt / cpu / Forward 0.000016775 s 0.000009523819971946069 s 1.76
slicing / JaXPipe / cpu / PreRev 0.000017621000000000003 s 0.000010216460013907636 s 1.72
slicing / JaXPipe / cpu / PostRev 0.000017052000000000002 s 0.000009857619970716768 s 1.73
slicing / JaXPipe / cpu / BothRev 0.000017267000000000003 s 0.000010476039997229235 s 1.65
slicing / Jax / cpu / BothRev 0.000017384 s 0.000009995319987865516 s 1.74
slicing / HLOOpt / cpu / PreRev 0.000017468 s 0.000010662620015864376 s 1.64
slicing / HLOOpt / cpu / PostRev 0.000017142 s 0.0000123771199832845 s 1.38
slicing / HLOOpt / cpu / BothRev 0.000017198 s 0.0000095578600121371 s 1.80
slicing / PartOpt / cpu / PreRev 0.000016913999999999998 s 0.000009616839970476575 s 1.76
slicing / PartOpt / cpu / PostRev 0.000017554 s 0.0000105017599707935 s 1.67
slicing / PartOpt / cpu / BothRev 0.000017604000000000003 s 0.000010213999994448386 s 1.72
slicing / IPartOpt / cpu / PreRev 0.000017507 s 0.000010010960013460136 s 1.75
slicing / IPartOpt / cpu / PostRev 0.000017063999999999998 s 0.000010388719992988626 s 1.64
slicing / IPartOpt / cpu / BothRev 0.000017242000000000002 s 0.000009856840006250422 s 1.75
slicing / DefOpt / cpu / PreRev 0.000017007 s 0.000009915399969031567 s 1.72
slicing / DefOpt / cpu / PostRev 0.000017169 s 0.000010333379996154693 s 1.66
slicing / DefOpt / cpu / BothRev 0.000017103 s 0.000010391340001660863 s 1.65
slicing / IDefOpt / cpu / PreRev 0.00001719 s 0.00000994755997453467 s 1.73
slicing / IDefOpt / cpu / PostRev 0.000017375000000000002 s 0.00000992942003904318 s 1.75
slicing / IDefOpt / cpu / BothRev 0.000017281999999999998 s 0.000009599200056982228 s 1.80
sum / JaXPipe / cpu / Primal 0.000007934459995340149 s 0.000007966920029502945 s 1.00
sum / Jax / cpu / Primal 0.000007713660013450862 s 0.000007496000016544712 s 1.03
sum / HLOOpt / cpu / Primal 0.000007801719993949518 s 0.0000075576799827103965 s 1.03
sum / PartOpt / cpu / Primal 0.0000076094599990028655 s 0.000007438460024786764 s 1.02
sum / IPartOpt / cpu / Primal 0.000008028899999317219 s 0.000008263000008810195 s 0.97
sum / DefOpt / cpu / Primal 0.000007553479988473554 s 0.000007974939999257913 s 0.95
sum / IDefOpt / cpu / Primal 0.000007622320006248627 s 0.000008394519963985659 s 0.91
sum / JaXPipe / cpu / Forward 0.000011628999993718026 s 0.00001140052003393066 s 1.02
sum / Jax / cpu / Forward 0.000011955720008245408 s 0.000011591580005188009 s 1.03
sum / HLOOpt / cpu / Forward 0.00001162217999535642 s 0.000011855700004161916 s 0.98
sum / PartOpt / cpu / Forward 0.000011605359998156929 s 0.000011221519962418824 s 1.03
sum / IPartOpt / cpu / Forward 0.000011456960000941765 s 0.000011349360020176392 s 1.01
sum / DefOpt / cpu / Forward 0.000011116159994344343 s 0.00001117800003157754 s 0.99
sum / IDefOpt / cpu / Forward 0.000011319420000290848 s 0.000011447939978097565 s 0.99
sum / JaXPipe / cpu / PreRev 0.000012075560000539554 s 0.000010888779997912936 s 1.11
sum / JaXPipe / cpu / PostRev 0.00001131110001097113 s 0.000010675600033209777 s 1.06
sum / JaXPipe / cpu / BothRev 0.00001103718000194931 s 0.00001096442001653486 s 1.01
sum / Jax / cpu / BothRev 0.000010945560004529398 s 0.000011158399984196876 s 0.98
sum / HLOOpt / cpu / PreRev 0.000011544879992015922 s 0.00001121798000895069 s 1.03
sum / HLOOpt / cpu / PostRev 0.00001297658000339652 s 0.000013005540022277274 s 1.00
sum / HLOOpt / cpu / BothRev 0.000010896380001668148 s 0.000011337859968989505 s 0.96
sum / PartOpt / cpu / PreRev 0.000010868199992728478 s 0.000011103740016551455 s 0.98
sum / PartOpt / cpu / PostRev 0.000011343680009758827 s 0.000010653500030457508 s 1.06
sum / PartOpt / cpu / BothRev 0.000011894620004113676 s 0.00001118933999350702 s 1.06
sum / IPartOpt / cpu / PreRev 0.000010643039995557049 s 0.000011035779980375082 s 0.96
sum / IPartOpt / cpu / PostRev 0.000010867799994684902 s 0.000011056240000471008 s 0.98
sum / IPartOpt / cpu / BothRev 0.00001098571999818887 s 0.000010846880031749608 s 1.01
sum / DefOpt / cpu / PreRev 0.000011377700009234104 s 0.000011112760003015865 s 1.02
sum / DefOpt / cpu / PostRev 0.000010743760001332702 s 0.000010988719977831352 s 0.98
sum / DefOpt / cpu / BothRev 0.000011137459996461984 s 0.000010721840008045548 s 1.04
sum / IDefOpt / cpu / PreRev 0.000010648100007983886 s 0.00001134413999352546 s 0.94
sum / IDefOpt / cpu / PostRev 0.000010931220006114015 s 0.000010876399992412189 s 1.01
sum / IDefOpt / cpu / BothRev 0.000011645060003502295 s 0.000011014359934051754 s 1.06
sum / JaXPipe / cuda / Primal 0.000002463 s 0.000002047 s 1.20
sum / Jax / cuda / Primal 0.000002464 s 0.000002048 s 1.20
sum / HLOOpt / cuda / Primal 0.000002464 s 0.000002047 s 1.20
sum / PartOpt / cuda / Primal 0.000002464 s 0.000002047 s 1.20
sum / IPartOpt / cuda / Primal 0.000002464 s 0.000002047 s 1.20
sum / DefOpt / cuda / Primal 0.000002464 s 0.000002048 s 1.20
sum / IDefOpt / cuda / Primal 0.000002463 s 0.000002047 s 1.20
sum / JaXPipe / cuda / Forward 0.000010304 s 0.000015263999999999998 s 0.68
sum / Jax / cuda / Forward 0.000010656 s 0.000010464 s 1.02
sum / HLOOpt / cuda / Forward 0.000010528 s 0.0000104 s 1.01
sum / PartOpt / cuda / Forward 0.000010368 s 0.000010304 s 1.01
sum / IPartOpt / cuda / Forward 0.000010624 s 0.000010272 s 1.03
sum / DefOpt / cuda / Forward 0.00001168 s 0.000010208 s 1.14
sum / IDefOpt / cuda / Forward 0.000011264 s 0.000010336 s 1.09
sum / JaXPipe / cuda / PreRev 0.000010272 s 0.000010048 s 1.02
sum / JaXPipe / cuda / PostRev 0.000010368 s 0.00001008 s 1.03
sum / JaXPipe / cuda / BothRev 0.000010336 s 0.000014112 s 0.73
sum / Jax / cuda / BothRev 0.00001056 s 0.0000096 s 1.10
sum / HLOOpt / cuda / PreRev 0.000010304 s 0.000009952 s 1.04
sum / HLOOpt / cuda / PostRev 0.000010432 s 0.000009633 s 1.08
sum / HLOOpt / cuda / BothRev 0.000010433 s 0.00001008 s 1.04
sum / PartOpt / cuda / PreRev 0.000010624 s 0.000010048 s 1.06
sum / PartOpt / cuda / PostRev 0.00001088 s 0.000009313 s 1.17
sum / PartOpt / cuda / BothRev 0.000010464 s 0.00000944 s 1.11
sum / IPartOpt / cuda / PreRev 0.00001024 s 0.000010047 s 1.02
sum / IPartOpt / cuda / PostRev 0.000010367 s 0.000009376 s 1.11
sum / IPartOpt / cuda / BothRev 0.000010335 s 0.000009984 s 1.04
sum / DefOpt / cuda / PreRev 0.000010144 s 0.00001008 s 1.01
sum / DefOpt / cuda / PostRev 0.000010143 s 0.00001008 s 1.01
sum / DefOpt / cuda / BothRev 0.000010208 s 0.000014433 s 0.71
sum / IDefOpt / cuda / PreRev 0.000010496 s 0.000010048 s 1.04
sum / IDefOpt / cuda / PostRev 0.000010144 s 0.000010112 s 1.00
sum / IDefOpt / cuda / BothRev 0.000010432 s 0.0000096 s 1.09
sum / JaXPipe / tpu / Primal 5.031e-7 s 5.103250000000001e-7 s 0.99
sum / Jax / tpu / Primal 5.469e-7 s 5.47525e-7 s 1.00
sum / HLOOpt / tpu / Primal 5.02975e-7 s 5.1055e-7 s 0.99
sum / PartOpt / tpu / Primal 5.468750000000001e-7 s 5.47175e-7 s 1.00
sum / IPartOpt / tpu / Primal 5.02925e-7 s 5.10875e-7 s 0.98
sum / DefOpt / tpu / Primal 5.47275e-7 s 5.472250000000001e-7 s 1.00
sum / IDefOpt / tpu / Primal 5.033499999999999e-7 s 5.10825e-7 s 0.99
sum / JaXPipe / tpu / Forward 0.0000015501999999999998 s 0.000001557325 s 1.00
sum / Jax / tpu / Forward 0.00000149295 s 0.0000014961749999999997 s 1.00
sum / HLOOpt / tpu / Forward 0.00000152835 s 0.000001532275 s 1.00
sum / PartOpt / tpu / Forward 0.000001488325 s 0.000001497125 s 0.99
sum / IPartOpt / tpu / Forward 0.000001528625 s 0.000001538 s 0.99
sum / DefOpt / tpu / Forward 0.000001497325 s 0.0000015041 s 1.00
sum / IDefOpt / tpu / Forward 0.000001546175 s 0.0000015345 s 1.01
sum / JaXPipe / tpu / PreRev 0.0000010184 s 0.000001003375 s 1.01
sum / JaXPipe / tpu / PostRev 0.000001066375 s 0.0000010455 s 1.02
sum / JaXPipe / tpu / BothRev 0.0000010263 s 9.98625e-7 s 1.03
sum / Jax / tpu / BothRev 0.000001069975 s 0.00000104185 s 1.03
sum / HLOOpt / tpu / PreRev 0.000001017525 s 0.00000101055 s 1.01
sum / HLOOpt / tpu / PostRev 0.00000106215 s 0.0000010539 s 1.01
sum / HLOOpt / tpu / BothRev 0.00000102965 s 0.000001004325 s 1.03
sum / PartOpt / tpu / PreRev 0.000001061625 s 0.0000010631 s 1.00
sum / PartOpt / tpu / PostRev 0.00000101875 s 0.00000101205 s 1.01
sum / PartOpt / tpu / BothRev 0.000001065725 s 0.000001043275 s 1.02
sum / IPartOpt / tpu / PreRev 0.000001021175 s 0.000001023525 s 1.00
sum / IPartOpt / tpu / PostRev 0.0000010638499999999998 s 0.000001059325 s 1.00
sum / IPartOpt / tpu / BothRev 0.000001022625 s 0.0000010066999999999998 s 1.02
sum / DefOpt / tpu / PreRev 0.0000010643250000000002 s 0.0000010547 s 1.01
sum / DefOpt / tpu / PostRev 0.0000010168499999999998 s 0.000001002725 s 1.01
sum / DefOpt / tpu / BothRev 0.0000010636 s 0.0000010368 s 1.03
sum / IDefOpt / tpu / PreRev 0.00000101695 s 0.000001004125 s 1.01
sum / IDefOpt / tpu / PostRev 0.0000010630000000000002 s 0.000001046575 s 1.02
sum / IDefOpt / tpu / BothRev 0.000001025275 s 0.000001000425 s 1.02
sum / JaXPipe / cpu / Primal 0.000014837 s 0.000007966920029502945 s 1.86
sum / Jax / cpu / Primal 0.000014667 s 0.000007496000016544712 s 1.96
sum / HLOOpt / cpu / Primal 0.000014715 s 0.0000075576799827103965 s 1.95
sum / PartOpt / cpu / Primal 0.000014829 s 0.000007438460024786764 s 1.99
sum / IPartOpt / cpu / Primal 0.00001465 s 0.000008263000008810195 s 1.77
sum / DefOpt / cpu / Primal 0.000014513 s 0.000007974939999257913 s 1.82
sum / IDefOpt / cpu / Primal 0.000014404 s 0.000008394519963985659 s 1.72
sum / JaXPipe / cpu / Forward 0.000019937 s 0.00001140052003393066 s 1.75
sum / Jax / cpu / Forward 0.000019874 s 0.000011591580005188009 s 1.71
sum / HLOOpt / cpu / Forward 0.000020025 s 0.000011855700004161916 s 1.69
sum / PartOpt / cpu / Forward 0.00002004 s 0.000011221519962418824 s 1.79
sum / IPartOpt / cpu / Forward 0.000019938 s 0.000011349360020176392 s 1.76
sum / DefOpt / cpu / Forward 0.00001967 s 0.00001117800003157754 s 1.76
sum / IDefOpt / cpu / Forward 0.00002012 s 0.000011447939978097565 s 1.76
sum / JaXPipe / cpu / PreRev 0.000018645 s 0.000010888779997912936 s 1.71
sum / JaXPipe / cpu / PostRev 0.000018461 s 0.000010675600033209777 s 1.73
sum / JaXPipe / cpu / BothRev 0.000018855 s 0.00001096442001653486 s 1.72
sum / Jax / cpu / BothRev 0.000018777 s 0.000011158399984196876 s 1.68
sum / HLOOpt / cpu / PreRev 0.000018887 s 0.00001121798000895069 s 1.68
sum / HLOOpt / cpu / PostRev 0.000019138 s 0.000013005540022277274 s 1.47
sum / HLOOpt / cpu / BothRev 0.000018674 s 0.000011337859968989505 s 1.65
sum / PartOpt / cpu / PreRev 0.000018538 s 0.000011103740016551455 s 1.67
sum / PartOpt / cpu / PostRev 0.000018847 s 0.000010653500030457508 s 1.77
sum / PartOpt / cpu / BothRev 0.000018537 s 0.00001118933999350702 s 1.66
sum / IPartOpt / cpu / PreRev 0.000018441 s 0.000011035779980375082 s 1.67
sum / IPartOpt / cpu / PostRev 0.000018821 s 0.000011056240000471008 s 1.70
sum / IPartOpt / cpu / BothRev 0.000018597 s 0.000010846880031749608 s 1.71
sum / DefOpt / cpu / PreRev 0.000018744 s 0.000011112760003015865 s 1.69
sum / DefOpt / cpu / PostRev 0.000018583 s 0.000010988719977831352 s 1.69
sum / DefOpt / cpu / BothRev 0.000018672 s 0.000010721840008045548 s 1.74
sum / IDefOpt / cpu / PreRev 0.000018749 s 0.00001134413999352546 s 1.65
sum / IDefOpt / cpu / PostRev 0.000018734 s 0.000010876399992412189 s 1.72
sum / IDefOpt / cpu / BothRev 0.000018614 s 0.000011014359934051754 s 1.69
value_and_grad / JaXPipe / cpu / Primal 0.000014669880006294988 s 0.00001462693999201292 s 1.00
value_and_grad / Jax / cpu / Primal 0.000014187800009040074 s 0.000013887280019844183 s 1.02
value_and_grad / HLOOpt / cpu / Primal 0.000013674479987457744 s 0.000013906800022596144 s 0.98
value_and_grad / PartOpt / cpu / Primal 0.000013841459997365746 s 0.00001392590001159988 s 0.99
value_and_grad / IPartOpt / cpu / Primal 0.000013940179994733626 s 0.000013735279972024728 s 1.01
value_and_grad / DefOpt / cpu / Primal 0.000013789780005026842 s 0.000014098959936745816 s 0.98
value_and_grad / IDefOpt / cpu / Primal 0.000013854460000857216 s 0.000014053879995117314 s 0.99
value_and_grad / JaXPipe / cuda / Primal 0.000032992 s 0.00003296 s 1.00
value_and_grad / Jax / cuda / Primal 0.000033152000000000004 s 0.000032800000000000004 s 1.01
value_and_grad / HLOOpt / cuda / Primal 0.000033088 s 0.000033472 s 0.99
value_and_grad / PartOpt / cuda / Primal 0.000033696 s 0.000033248 s 1.01
value_and_grad / IPartOpt / cuda / Primal 0.000033087 s 0.000033152000000000004 s 1.00
value_and_grad / DefOpt / cuda / Primal 0.000033312 s 0.000033408 s 1.00
value_and_grad / IDefOpt / cuda / Primal 0.000033504 s 0.000034176 s 0.98
value_and_grad / JaXPipe / tpu / Primal 0 s 0 s 1
value_and_grad / Jax / tpu / Primal 0 s 0 s 1
value_and_grad / HLOOpt / tpu / Primal 0 s 0 s 1
value_and_grad / PartOpt / tpu / Primal 0 s 0 s 1
value_and_grad / IPartOpt / tpu / Primal 0 s 0 s 1
value_and_grad / DefOpt / tpu / Primal 0 s 0 s 1
value_and_grad / IDefOpt / tpu / Primal 0 s 0 s 1
value_and_grad / JaXPipe / cpu / Primal 0.000023072 s 0.00001462693999201292 s 1.58
value_and_grad / Jax / cpu / Primal 0.000022711 s 0.000013887280019844183 s 1.64
value_and_grad / HLOOpt / cpu / Primal 0.000022896 s 0.000013906800022596144 s 1.65
value_and_grad / PartOpt / cpu / Primal 0.000022783 s 0.00001392590001159988 s 1.64
value_and_grad / IPartOpt / cpu / Primal 0.000022572 s 0.000013735279972024728 s 1.64
value_and_grad / DefOpt / cpu / Primal 0.000022882 s 0.000014098959936745816 s 1.62
value_and_grad / IDefOpt / cpu / Primal 0.000022605 s 0.000014053879995117314 s 1.61
jaxmd20 / JaXPipe / cuda / Primal 0.001488856 s 0.00146413 s 1.02
jaxmd20 / Jax / cuda / Primal 0.001456056 s 0.001500002 s 0.97
jaxmd20 / HLOOpt / cuda / Primal 0.001328184 s 0.001340257 s 0.99
jaxmd20 / PartOpt / cuda / Primal 0.001373625 s 0.001326465 s 1.04
jaxmd20 / IPartOpt / cuda / Primal 0.001367481 s 0.001347746 s 1.01
jaxmd20 / DefOpt / cuda / Primal 0.000947771 s 0.0009191679999999 s 1.03
jaxmd20 / IDefOpt / cuda / Primal 0.00098441 s 0.000950881 s 1.04
jaxmd20 / JaXPipe / cuda / Forward 0.001634583 s 0.001554785 s 1.05
jaxmd20 / Jax / cuda / Forward 0.001868471 s 0.0017800659999999 s 1.05
jaxmd20 / HLOOpt / cuda / Forward 0.001724215 s 0.001616514 s 1.07
jaxmd20 / PartOpt / cuda / Forward 0.001746135 s 0.001637218 s 1.07
jaxmd20 / IPartOpt / cuda / Forward 0.00171135 s 0.001613826 s 1.06
jaxmd20 / DefOpt / cuda / Forward 0.001724568 s 0.0016381149999999 s 1.05
jaxmd20 / IDefOpt / cuda / Forward 0.001732119 s 0.001622914 s 1.07
jaxmd20 / JaXPipe / cuda / PreRev 0.002758928 s 0.002663619 s 1.04
jaxmd20 / JaXPipe / cuda / PostRev 0.005434434 s 0.005329255 s 1.02
jaxmd20 / JaXPipe / cuda / BothRev 0.002740529 s 0.002686564 s 1.02
jaxmd20 / Jax / cuda / BothRev 0.005435619 s 0.005338087 s 1.02
jaxmd20 / HLOOpt / cuda / PreRev 0.002844401 s 0.002748932 s 1.03
jaxmd20 / HLOOpt / cuda / PostRev 0.005481699 s 0.005346055 s 1.03
jaxmd20 / HLOOpt / cuda / BothRev 0.002822161 s 0.002751748 s 1.03
jaxmd20 / PartOpt / cuda / PreRev 0.002899473 s 0.00281338 s 1.03
jaxmd20 / PartOpt / cuda / PostRev 0.0056148499999999 s 0.0053920389999999 s 1.04
jaxmd20 / PartOpt / cuda / BothRev 0.002835634 s 0.002791875 s 1.02
jaxmd20 / IPartOpt / cuda / PreRev 0.002940144 s 0.0028067549999999 s 1.05
jaxmd20 / IPartOpt / cuda / PostRev 0.0055639709999999 s 0.005378982 s 1.03
jaxmd20 / IPartOpt / cuda / BothRev 0.002865584 s 0.002749955 s 1.04
jaxmd20 / DefOpt / cuda / PreRev 0.002902353 s 0.0028253149999999 s 1.03
jaxmd20 / DefOpt / cuda / PostRev 0.002883792 s 0.002760165 s 1.04
jaxmd20 / DefOpt / cuda / BothRev 0.0028239849999999 s 0.0027866279999999 s 1.01
jaxmd20 / IDefOpt / cuda / PreRev 0.002926224 s 0.002807364 s 1.04
jaxmd20 / IDefOpt / cuda / PostRev 0.002352115 s 0.002303811 s 1.02
jaxmd20 / IDefOpt / cuda / BothRev 0.0028343189999999 s 0.0027489629999999 s 1.03
jaxmd20 / JaXPipe / tpu / Primal 0.00927588125 s 0.0092963131249999 s 1.00
jaxmd20 / Jax / tpu / Primal 0.0092636749999999 s 0.00927079625 s 1.00
jaxmd20 / HLOOpt / tpu / Primal 0.0091532943749999 s 0.009157256875 s 1.00
jaxmd20 / PartOpt / tpu / Primal 0.009195653125 s 0.00919751875 s 1.00
jaxmd20 / IPartOpt / tpu / Primal 0.009201291875 s 0.009203176875 s 1.00
jaxmd20 / DefOpt / tpu / Primal 0.008793926875 s 0.008796180625 s 1.00
jaxmd20 / IDefOpt / tpu / Primal 0.008698733125 s 0.008703285625 s 1.00
jaxmd20 / JaXPipe / tpu / Forward 0.01742005625 s 0.0174139475 s 1.00
jaxmd20 / Jax / tpu / Forward 0.01872754 s 0.0187393075 s 1.00
jaxmd20 / HLOOpt / tpu / Forward 0.01739210625 s 0.0174003049999999 s 1.00
jaxmd20 / PartOpt / tpu / Forward 0.017412095625 s 0.017413809375 s 1.00
jaxmd20 / IPartOpt / tpu / Forward 0.01741418 s 0.0174123799999999 s 1.00
jaxmd20 / DefOpt / tpu / Forward 0.01741347 s 0.017426431875 s 1.00
jaxmd20 / IDefOpt / tpu / Forward 0.017413868125 s 0.017412701875 s 1.00
jaxmd20 / JaXPipe / tpu / PreRev 0.02545243625 s 0.0254493456249999 s 1.00
jaxmd20 / JaXPipe / tpu / PostRev 0.0218912912499999 s 0.021875376875 s 1.00
jaxmd20 / JaXPipe / tpu / BothRev 0.02547337375 s 0.025473811875 s 1.00
jaxmd20 / Jax / tpu / BothRev 0.02188864875 s 0.021873763125 s 1.00
jaxmd20 / HLOOpt / tpu / PreRev 0.02558511 s 0.02558455 s 1.00
jaxmd20 / HLOOpt / tpu / PostRev 0.020733736875 s 0.020714709375 s 1.00
jaxmd20 / HLOOpt / tpu / BothRev 0.02568037125 s 0.025694736875 s 1.00
jaxmd20 / PartOpt / tpu / PreRev 0.02550273125 s 0.025487941875 s 1.00
jaxmd20 / PartOpt / tpu / PostRev 0.0215117274999999 s 0.02153581625 s 1.00
jaxmd20 / PartOpt / tpu / BothRev 0.0255963625 s 0.0255683875 s 1.00
jaxmd20 / IPartOpt / tpu / PreRev 0.0254764181249999 s 0.0254793275 s 1.00
jaxmd20 / IPartOpt / tpu / PostRev 0.021537655 s 0.0215210425 s 1.00
jaxmd20 / IPartOpt / tpu / BothRev 0.025551655625 s 0.02557064375 s 1.00
jaxmd20 / DefOpt / tpu / PreRev 0.0255062624999999 s 0.0254876425 s 1.00
jaxmd20 / DefOpt / tpu / PostRev 0.01880501625 s 0.018822664375 s 1.00
jaxmd20 / DefOpt / tpu / BothRev 0.025596939375 s 0.02557184625 s 1.00
jaxmd20 / IDefOpt / tpu / PreRev 0.025478773125 s 0.025477181875 s 1.00
jaxmd20 / IDefOpt / tpu / PostRev 0.01834934375 s 0.018344374375 s 1.00
jaxmd20 / IDefOpt / tpu / BothRev 0.0255535225 s 0.0255753725 s 1.00
jaxmd40 / JaXPipe / cpu / Primal 0.068946917 s 0.0755081639999999 s 0.91
jaxmd40 / Jax / cpu / Primal 0.064149185 s 0.0717200289999999 s 0.89
jaxmd40 / HLOOpt / cpu / Primal 0.089912346 s 0.108675883 s 0.83
jaxmd40 / PartOpt / cpu / Primal 0.061760006 s 0.082305401 s 0.75
jaxmd40 / IPartOpt / cpu / Primal 0.0620066109999999 s 0.084469558 s 0.73
jaxmd40 / DefOpt / cpu / Primal 0.08883877 s 0.10659383 s 0.83
jaxmd40 / IDefOpt / cpu / Primal 0.08587914 s 0.111599474 s 0.77
jaxmd40 / JaXPipe / cpu / Forward 0.162879676 s 0.189724921 s 0.86
jaxmd40 / Jax / cpu / Forward 0.085562472 s 0.089788619 s 0.95
jaxmd40 / HLOOpt / cpu / Forward 0.155593933 s 0.188932056 s 0.82
jaxmd40 / PartOpt / cpu / Forward 0.1510307709999999 s 0.187766515 s 0.80
jaxmd40 / IPartOpt / cpu / Forward 0.154250584 s 0.194974803 s 0.79
jaxmd40 / DefOpt / cpu / Forward 0.155667063 s 0.184983297 s 0.84
jaxmd40 / IDefOpt / cpu / Forward 0.154128999 s 0.190125706 s 0.81
jaxmd40 / JaXPipe / cpu / PreRev 0.222298178 s 0.251550023 s 0.88
jaxmd40 / JaXPipe / cpu / PostRev 0.133741441 s 0.155112812 s 0.86
jaxmd40 / JaXPipe / cpu / BothRev 0.2146169319999999 s 0.248999614 s 0.86
jaxmd40 / Jax / cpu / BothRev 0.134790142 s 0.162482192 s 0.83
jaxmd40 / HLOOpt / cpu / PreRev 0.211584163 s 0.248515938 s 0.85
jaxmd40 / HLOOpt / cpu / PostRev 0.167231734 s 0.213123876 s 0.78
jaxmd40 / HLOOpt / cpu / BothRev 0.239498848 s 0.290243502 s 0.83
jaxmd40 / PartOpt / cpu / PreRev 0.217190847 s 0.243762015 s 0.89
jaxmd40 / PartOpt / cpu / PostRev 0.128042202 s 0.145630436 s 0.88
jaxmd40 / PartOpt / cpu / BothRev 0.224680196 s 0.267546272 s 0.84
jaxmd40 / IPartOpt / cpu / PreRev 0.2093421329999999 s 0.246754119 s 0.85
jaxmd40 / IPartOpt / cpu / PostRev 0.130794342 s 0.134807949 s 0.97
jaxmd40 / IPartOpt / cpu / BothRev 0.245755313 s 0.2652395629999999 s 0.93
jaxmd40 / DefOpt / cpu / PreRev 0.233444618 s 0.255335409 s 0.91
jaxmd40 / DefOpt / cpu / PostRev 0.176763735 s 0.216085604 s 0.82
jaxmd40 / DefOpt / cpu / BothRev 0.248300952 s 0.287981373 s 0.86
jaxmd40 / IDefOpt / cpu / PreRev 0.207835264 s 0.249678795 s 0.83
jaxmd40 / IDefOpt / cpu / PostRev 0.168031569 s 0.210582048 s 0.80
jaxmd40 / IDefOpt / cpu / BothRev 0.252312424 s 0.26354349 s 0.96
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / cuda / Primal 1.700558847 s 1.703403109 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / cuda / Primal 1.703825022 s 1.704959474 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / cuda / Primal 1.715774473 s 1.714279019 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / cuda / Primal 1.695148232 s 1.695521465 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / cuda / Primal 1.693305238 s 1.694030017 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / cuda / Primal 1.663501122 s 1.664191437 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / cuda / Primal 1.912728565 s 1.914000938 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / tpu / Primal 3.038800781875 s 3.038906811875 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / tpu / Primal 3.0394918675 s 3.03936133 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / tpu / Primal 3.12158328625 s 3.12152741125 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / tpu / Primal 3.060350193125 s 3.060072164375 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / tpu / Primal 3.0605353625 s 3.060387236875 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / tpu / Primal 2.1024384975 s 2.10243730625 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / tpu / Primal 2.94827600125 s 2.94835662125 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / JaXPipe / cpu / Primal 5.908700971999999 s 6.770451712 s 0.87
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / Jax / cpu / Primal 5.795272419 s 6.796283552 s 0.85
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / HLOOpt / cpu / Primal 5.839469013 s 6.784701461 s 0.86
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / PartOpt / cpu / Primal 5.901418588 s 6.773146511 s 0.87
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IPartOpt / cpu / Primal 5.984600129 s 6.831082725 s 0.88
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / DefOpt / cpu / Primal 2.223750266 s 2.680127355 s 0.83
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IDefOpt / cpu / Primal 6.305866134 s 7.33637517 s 0.86

This comment was automatically generated by workflow using github-action-benchmark.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants