[WebGPU] Optimize GEMM with vec4

[xiaofeihan1]

Description

In this PR, we use vec4 to optimize GEMM when colums of A and B can be divided by 4, or use previous shader.
I will add u32/vec2 implementation in the future, and we will only keep one shader at that time.

Perf comparison

I run customized model only include GEMM(M = N = K = 1024) with nodejs on M2/M3 Max. Roughly 20% increase.

	!transA&&!transB	transA	transB	transA&&transB
M2	9.36->7.41	9.45->7.54	11.21->8.19	9.66->8.37
M3 max	8.07->6.99	7.54->6.53	8.42->5.89	5.47->5.29

[fs-eire]

Is there a way to reuse the implementation of MatMul? My understanding is that there are some kind of duplication between GEMM and MatMul, and it would be great if we can reuse the shared code

[xiaofeihan1]

Is there a way to reuse the implementation of MatMul? My understanding is that there are some kind of duplication between GEMM and MatMul, and it would be great if we can reuse the shared code

Thanks for the callout. That's what I'm gonna do next.
For current PR, I want to push forward to support vec4 for GEMM. I will take the refactor work in future PRs because it also require some effort to consider(e.g.There are also some differences between gemm and matmul, e.g. the latter supports batch size, the former supports transpose, etc). WDYT?

[fs-eire]

/azp run Linux QNN CI Pipeline, Win_TRT_Minimal_CUDA_Test_CI, Windows ARM64 QNN CI Pipeline, Windows GPU Doc Gen CI Pipeline, Windows x64 QNN CI Pipeline

[azure-pipelines]

Azure Pipelines successfully started running 5 pipeline(s).

[fs-eire]

/azp run Linux QNN CI Pipeline, Win_TRT_Minimal_CUDA_Test_CI, Windows ARM64 QNN CI Pipeline, Windows GPU Doc Gen CI Pipeline, Windows x64 QNN CI Pipeline

[azure-pipelines]

Azure Pipelines successfully started running 5 pipeline(s).

[qjia7]

LGTM, thanks.

[fs-eire]

/azp run Windows ARM64 QNN CI Pipeline,Windows x64 QNN CI Pipeline,Windows GPU Doc Gen CI Pipeline,ONNX Runtime Web CI Pipeline,Win_TRT_Minimal_CUDA_Test_CI,Linux CPU CI Pipeline,Linux CPU Minimal Build E2E CI Pipeline,Linux GPU CI Pipeline,Linux GPU TensorRT CI Pipeline,Linux OpenVINO CI Pipeline

[fs-eire]

/azp run Linux QNN CI Pipeline,onnxruntime-binary-size-checks-ci-pipeline,Big Models,Linux Android Emulator QNN CI Pipeline,Android CI Pipeline,iOS CI Pipeline,ONNX Runtime React Native CI Pipeline,Linux DNNL CI Pipeline,Linux MIGraphX CI Pipeline,Linux ROCm CI Pipeline

[azure-pipelines]

Azure Pipelines successfully started running 5 pipeline(s).

[azure-pipelines]

Azure Pipelines successfully started running 6 pipeline(s), but failed to run 1 pipeline(s).