UPSTREAM PR #20635: [CUDA] Increase number of output elements per-thread block if the K-dimension is small#1275
Open
UPSTREAM PR #20635: [CUDA] Increase number of output elements per-thread block if the K-dimension is small#1275
Conversation
With tensor parallelism, the K-dimension of the FFN-down matrices is split, which makes it quite small, especially for MOEs. For example, Qwen3-30b-A3B has a K-dimension of 768, and Qwen3235B-A22B has k-dimension of 1536. The current heuristic uses a group of 4 warps irrespective of K-dimension size, resulting in some of the threads being idle. This results in poor performance for these matrices. This change increases the number of output elements per block for such cases.
|
No meaningful performance changes were detected across 120772 analyzed functions in the following binaries: build.bin.libllama.so, build.bin.llama-bench, build.bin.llama-cvector-generator, build.bin.llama-tts, build.bin.libmtmd.so, build.bin.libggml-base.so, build.bin.libggml-cpu.so, build.bin.libggml.so, build.bin.llama-gguf-split, build.bin.llama-llava-cli, build.bin.llama-minicpmv-cli, build.bin.llama-quantize, build.bin.llama-qwen2vl-cli, build.bin.llama-tokenize, build.bin.llama-gemma3-cli. 🔎 Full breakdown: Loci Inspector |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Note
Source pull request: ggml-org/llama.cpp#20635
The K-dimension (inner dot product dimension) of the FFN-down matrices can be quite small, especially for MOEs. For example, Qwen3-30b-A3B has a K-dimension of 768, and Qwen3-235B-A22B has a k-dimension of 1536. The current heuristic uses a group of 4 warps irrespective of K-dimension size, resulting in some of the threads being idle. This results in poor performance for these matrices.
This change increases the number of output elements per block for such matrices.
This change is also helpful for Tensor parallelism (PR ggml-org/llama.cpp#19378), where FFN-down is split along the K dimension.
Single GPU Performance on 1x RTX Pro 6000 Blackwell
Tensor Parallelism Performance on 2x RTX Pro 6000 Blackwell with PR 19378