Open
Conversation
GPU-accelerate two proving bottlenecks via Apple Metal: 1. Merkle tree construction (Poseidon2) — GPU hashing for trees with 2^13 to 2^20 leaves 2. Quotient polynomial evaluation — fused gate eval + alpha reduction with zero device memory allocation Enable with `features = ["metal"]`. Automatic runtime dispatch via TypeId checks; falls back to CPU when conditions are not met. Benchmark (lighter-prover, 500 txs, 125 chunks): - CPU-only: 653.5s - GPU Metal: 471.6s (1.39x, 27.8% faster) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Verified with back-to-back runs: speedup ranges 21-28% across runs due to thermal/load variance. Report the conservative 1.27x (21.3%). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Shows Apple M4 GPU History: active with Metal on, idle with Metal off. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Shows Merkle vs Quotient contribution for each circuit type. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Remove per-gate Instant::now() + AtomicU64 profiling that broke compiler auto-vectorization and added ~40% overhead to CPU gate evaluation. GPU quotient kernel and caching remain enabled. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- GPU_BENCHMARK_COMBINED.md: Metal vs CUDA side-by-side comparison - GPU_COST_ANALYSIS.md: AWS cost per proof analysis (mac-m4 vs g6e L40S) - METAL_BENCH_GUIDE.md: Step-by-step guide for running Metal benchmark - cost_per_proof.png + plot_cost.py: Cost comparison bar chart - GPU_BENCHMARK_RESULTS.md: Updated with validated 1.56-1.61x speedup numbers Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
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
2 participants
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.
No description provided.