Repository navigation
Lower concrete BF16 GEMM to Grider ISA - #16
Merged
Merged
Conversation
SaibaWaipu
changed the title
SaibaWaipu
changed the base branch from
codex/issue-8-bf16-f32-accumulator
to
main
October 30, 2026 09:27
SaibaWaipu
marked this pull request as ready for review
October 30, 2026 09:27
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Depends on #11 (FP32 BF16 MAC semantics). Adds a concrete 16x16 BF16 GEMM lowering using row/column lane masks, immediate loads, SIMD MACs, and final BF16 conversion. It includes a Python execution report (instruction count, cycle estimate, maximum error) and a utility to run the generated stream through the tile RTL testbench.
Scope
The current lowering embeds matrices in
MOVIimmediates because the ISA has no runtime global-memory load or program fetch. This is a complete fixed-input lowering path, not yet a reusable runtime GEMM kernel. Draft pending RTL/Python execution and reference-error comparison. No tests were run.