[Kernel] Use CUTLASS kernels for the FP8 layers with Bias

[tlrmchlsmth]

FP8 biases work as of #5931

I am seeing speedup on Qwen2-7b, which has biases: https://docs.google.com/spreadsheets/d/1faYp80qbmOra3iRMHgga7dPiC6tWULpPQHN_5BfM7gs/

---

PR Checklist (Click to Expand)

Thank you for your contribution to vLLM! Before submitting the pull request, please ensure the PR meets the following criteria. This helps vLLM maintain the code quality and improve the efficiency of the review process.

PR Title and Classification

Only specific types of PRs will be reviewed. The PR title is prefixed appropriately to indicate the type of change. Please use one of the following:

• [Bugfix] for bug fixes.
• [CI/Build] for build or continuous integration improvements.
• [Doc] for documentation fixes and improvements.
• [Model] for adding a new model or improving an existing model. Model name should appear in the title.
• [Frontend] For changes on the vLLM frontend (e.g., OpenAI API server, LLM class, etc.)
• [Kernel] for changes affecting CUDA kernels or other compute kernels.
• [Core] for changes in the core vLLM logic (e.g., LLMEngine, AsyncLLMEngine, Scheduler, etc.)
• [Hardware][Vendor] for hardware-specific changes. Vendor name should appear in the prefix (e.g., [Hardware][AMD]).
• [Misc] for PRs that do not fit the above categories. Please use this sparingly.

Note: If the PR spans more than one category, please include all relevant prefixes.

Code Quality

The PR need to meet the following code quality standards:

• We adhere to Google Python style guide and Google C++ style guide.
• Pass all linter checks. Please use format.sh to format your code.
• The code need to be well-documented to ensure future contributors can easily understand the code.
• Include sufficient tests to ensure the project to stay correct and robust. This includes both unit tests and integration tests.
• Please add documentation to docs/source/ if the PR modifies the user-facing behaviors of vLLM. It helps vLLM user understand and utilize the new features or changes.

Notes for Large Changes

Please keep the changes as concise as possible. For major architectural changes (>500 LOC excluding kernel/data/config/test), we would expect a GitHub issue (RFC) discussing the technical design and justification. Otherwise, we will tag it with rfc-required and might not go through the PR.

What to Expect for the Reviews

The goal of the vLLM team is to be a transparent reviewing machine. We would like to make the review process transparent and efficient and make sure no contributor feel confused or frustrated. However, the vLLM team is small, so we need to prioritize some PRs over others. Here is what you can expect from the review process:

• After the PR is submitted, the PR will be assigned to a reviewer. Every reviewer will pick up the PRs based on their expertise and availability.
• After the PR is assigned, the reviewer will provide status update every 2-3 days. If the PR is not reviewed within 7 days, please feel free to ping the reviewer or the vLLM team.
• After the review, the reviewer will put an action-required label on the PR if there are changes required. The contributor should address the comments and ping the reviewer to re-review the PR.
• Please respond to all comments within a reasonable time frame. If a comment isn't clear or you disagree with a suggestion, feel free to ask for clarification or discuss the suggestion.

Thank You

Finally, thank you for taking the time to read these guidelines and for your interest in contributing to vLLM. Your contributions make vLLM a great tool for everyone!

[tlrmchlsmth]

@cyang49 were you seeing higher performance with the CUTLASS kernels on Granite as well?

[cyang49]

@cyang49 were you seeing higher performance with the CUTLASS kernels on Granite as well?

Yes. There was a slight increase in throughput from my tests.

BTW, I recall that even though the 4 linear layers of the transformer block all have bias, in vllm only 2 of them will go through the path where the bias add is fused. @tdoublep looked into that logic and may be able to provide more details.

[tlrmchlsmth]

@cyang49 Is it because of the code in RowParallelLinear here?

vllm/vllm/model_executor/layers/linear.py

Lines 755 to 769 in b422d49

	# Matrix multiply.
	assert self.quant_method is not None
	output_parallel = self.quant_method.apply(self, input_parallel)
	if self.reduce_results and self.tp_size > 1:
	output_ = tensor_model_parallel_all_reduce(output_parallel)
	else:
	output_ = output_parallel
	if not self.skip_bias_add:
	output = output_ + self.bias if self.bias is not None else output_
	output_bias = None
	else:
	output = output_
	output_bias = self.bias
	return output, output_bias

I don't know why it's like this, but we can look into fixing that.

[tdoublep]

@tlrmchlsmth yeah it's exactly that code (which I guess is somehow related to TP>1).

[tdoublep]

I think we should do this at least: #6327