Add cutlass decode kernel to TritonBench #18798

Job	Run time
build-macos (macos-13, static)	12m 49s
build-bazel (x86, linux.12xlarge, clang)	9m 23s
build-bazel (x86, linux.12xlarge, gcc)	10m 8s
build-linux (x86, linux.12xlarge, static, clang, 16.0.6)	20m 13s
build-linux (x86, linux.12xlarge, static, gcc, 14.1.0)	26m 21s
build-linux (x86, linux.12xlarge, static, gcc, 9.5.0)	26m 0s
build-macos (macos-13, shared)	17m 27s
build-linux (arm, linux.arm64.m7g.4xlarge, static, gcc, 14.1.0)	1h 26m 57s
build-linux (x86, linux.12xlarge, shared, gcc, 14.1.0)	26m 27s
build-linux (x86, linux.12xlarge, shared, clang, 16.0.6)	19m 54s
build-linux (arm, linux.arm64.m7g.4xlarge, shared, clang, 16.0.6)	32m 51s
build-linux (arm, linux.arm64.m7g.4xlarge, static, clang, 16.0.6)	33m 9s
build-windows (windows-2022, shared)	12m 32s
build-linux (x86, linux.12xlarge, shared, gcc, 9.5.0)	25m 42s
build-linux (arm, linux.arm64.m7g.4xlarge, shared, gcc, 14.1.0)	1h 27m 27s
build-windows (windows-2022, static)	12m 7s
	7h 39m 27s

Provide feedback