nki(quantize_global): NKI (Trainium) implementation - #295
Open
bowencui123 wants to merge 6 commits into
Open
bowencui123 wants to merge 6 commits into
bowencui123 wants to merge 6 commits into
Conversation
Split out of the consolidated NKI branch cecilia/feature/nki-vector-add (nki-all-operators, PR #259) so each operator can be reviewed on its own. Co-Authored-By: Cecilia123li <68335867+Cecilia123li@users.noreply.github.com> Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012Q38kGmXvyoeM1qtCbheSL
Tunables mirror the Triton search space (`BLOCK_SIZE`); defaults are the previous constants, so autotune=False is unchanged. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012Q38kGmXvyoeM1qtCbheSL
…NC2) Merges NKI backend timing into results/csv/quantize_global_default.csv, run against this branch's impl_nki.py on trn2.3xlarge with the LNC2 execution contract (NEURON_LOGICAL_NC_CONFIG=2, NEURON_RT_NUM_CORES=1, NEURON_CC_FLAGS="--target trn2 --lnc 2"). Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01AQseF7nyesBh8KZAp8g7Cm
…e, LNC2) Merges NKI backend timing into results/csv/quantize_global_autotune.csv, run with --autotune against this branch's impl_nki.py on trn2.3xlarge, LNC2 execution contract. All cases pass correctness verification. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01AQseF7nyesBh8KZAp8g7Cm
…obal # Conflicts: # results/csv/quantize_global_autotune.csv # results/csv/quantize_global_default.csv
…pad/reshape); rerun default+autotune benchmarks Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ScXYNjrrKGgDUVNHxv7HJt
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
NKI (AWS Trainium) implementation of quantize_global, split out of the consolidated NKI branch
cecilia/feature/nki-vector-add(nki-all-operators, #259) so each operator can be reviewed independently.Files: A benchmarks/operators/quantize_global/impl_nki.py
Status: imports and exposes run()/get_last_config() on trn2 (nki 0.6.0); not individually re-benchmarked in this split
Implementation by @Cecilia123li. Timing/identity infrastructure: #261; Trainium peak/roofline infra: #262.
🤖 Generated with Claude Code
https://claude.ai/code/session_012Q38kGmXvyoeM1qtCbheSL
Autotune (c2d339a)
block_size(free-dim tile per 128-partition block)BLOCK_SIZEautotune=Falsekeeps the previous constants (default numbers unchanged). Validation on trn2, case 0 (default run + autotune code path with the candidate timer stubbed — no sweep;--autotuneruns a real sweep):