Replacing L1 base address increment instructions with CFGSHIFTMASK #17723
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
atatuzunerTT
force-pushed
the
atuzuner/CFGSHIFTMASK
branch
2 times, most recently
from
24313ee to
3bc159f
Compare
atatuzunerTT
requested review from
TT-BrianLiu,
bbradelTT,
yugaoTT,
vsureshTT,
edwinleeTT,
rtawfik01,
ttmtrajkovic and
rdjogoTT
as code owners
rtawfik01
approved these changes
Feb 19, 2025
bbradelTT
approved these changes
Feb 19, 2025
atatuzunerTT
force-pushed
the
atuzuner/CFGSHIFTMASK
branch
from
3bc159f to
5d6096e
Compare
dgomezTT
pushed a commit
that referenced
this pull request
Feb 19, 2025
…17723) ### Ticket [Link to Github Issue](tenstorrent/tt-llk-bh#4) ### Problem description Blackhole has new `CFGSHIFTMASK` that can update addresses for the unpacker instructions inside the mop/replay buffers. If an operation is unpacker bound, then using this instruction should increase performance. ### What's changed Replaced L1 base address increment code that uses cfg read/write and tdma gpr operations with the new `CFGSHIFTMASK` instruction in the unpack AB matmul llk api. This replacement saves 6 instructions in the mop replay buffer. No notable performance improvements. Only affects BH and addresses an issue in BH third party repo. ### Checklist - [x] [All post commit](https://github.com/tenstorrent/tt-metal/actions/workflows/all-post-commit-workflows.yaml) [CI passes](https://github.com/tenstorrent/tt-metal/actions/runs/13399863311) - [x] [Blackhole Post commit](https://github.com/tenstorrent/tt-metal/actions/workflows/blackhole-post-commit.yaml) [CI passes](https://github.com/tenstorrent/tt-metal/actions/runs/13399865409) (if applicable) - [ ] [Model regression](https://github.com/tenstorrent/tt-metal/actions/workflows/perf-models.yaml) CI passes (if applicable) - [ ] [Device performance regression](https://github.com/tenstorrent/tt-metal/actions/workflows/perf-device-models.yaml) CI passes (if applicable) - [ ] **(For models and ops writers)** Full [new models tests](https://github.com/tenstorrent/tt-metal/actions/workflows/full-new-models-suite.yaml) CI passes (if applicable) - [ ] New/Existing tests provide coverage for changes
hschoi4448
pushed a commit
that referenced
this pull request
Feb 20, 2025
…17723) ### Ticket [Link to Github Issue](tenstorrent/tt-llk-bh#4) ### Problem description Blackhole has new `CFGSHIFTMASK` that can update addresses for the unpacker instructions inside the mop/replay buffers. If an operation is unpacker bound, then using this instruction should increase performance. ### What's changed Replaced L1 base address increment code that uses cfg read/write and tdma gpr operations with the new `CFGSHIFTMASK` instruction in the unpack AB matmul llk api. This replacement saves 6 instructions in the mop replay buffer. No notable performance improvements. Only affects BH and addresses an issue in BH third party repo. ### Checklist - [x] [All post commit](https://github.com/tenstorrent/tt-metal/actions/workflows/all-post-commit-workflows.yaml) [CI passes](https://github.com/tenstorrent/tt-metal/actions/runs/13399863311) - [x] [Blackhole Post commit](https://github.com/tenstorrent/tt-metal/actions/workflows/blackhole-post-commit.yaml) [CI passes](https://github.com/tenstorrent/tt-metal/actions/runs/13399865409) (if applicable) - [ ] [Model regression](https://github.com/tenstorrent/tt-metal/actions/workflows/perf-models.yaml) CI passes (if applicable) - [ ] [Device performance regression](https://github.com/tenstorrent/tt-metal/actions/workflows/perf-device-models.yaml) CI passes (if applicable) - [ ] **(For models and ops writers)** Full [new models tests](https://github.com/tenstorrent/tt-metal/actions/workflows/full-new-models-suite.yaml) CI passes (if applicable) - [ ] New/Existing tests provide coverage for changes
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Ticket
Link to Github Issue
Problem description
Blackhole has new
CFGSHIFTMASKthat can update addresses for the unpacker instructions inside the mop/replay buffers. If an operation is unpacker bound, then using this instruction should increase performance.What's changed
Replaced L1 base address increment code that uses cfg read/write and tdma gpr operations with the new
CFGSHIFTMASKinstruction in the unpack AB matmul llk api. This replacement saves 6 instructions in the mop replay buffer. No notable performance improvements.Only affects BH and addresses an issue in BH third party repo.
Checklist