mirror of
https://github.com/tile-ai/tilelang.git
synced 2026-10-21 06:40:23 +08:00
Lei Wang
99ee74b343
* Support int4 T.gemm and direct packed cp.async lowering * Keep cp.async vectorization as a late fallback * Change tl.ptx_cp_async to use num_elems semantics * fix test
41 lines
1.3 KiB
Python
41 lines
1.3 KiB
Python
import tilelang
|
|
import tilelang.testing
|
|
import tilelang.language as T
|
|
|
|
|
|
def matmul_nt_int4(M, N, K, block_M, block_N, block_K):
|
|
@T.prim_func
|
|
def main(
|
|
A: T.Tensor((M, K), T.int4),
|
|
B: T.Tensor((N, K), T.int4),
|
|
C: T.Tensor((M, N), T.int32),
|
|
):
|
|
with T.Kernel(T.ceildiv(N, block_N), T.ceildiv(M, block_M), threads=128) as (bx, by):
|
|
A_shared = T.alloc_shared((block_M, block_K), T.int4)
|
|
B_shared = T.alloc_shared((block_N, block_K), T.int4)
|
|
C_local = T.alloc_fragment((block_M, block_N), T.int32)
|
|
|
|
T.clear(C_local)
|
|
for ko in T.Pipelined(T.ceildiv(K, block_K), num_stages=3):
|
|
T.copy(A[by * block_M, ko * block_K], A_shared)
|
|
T.copy(B[bx * block_N, ko * block_K], B_shared)
|
|
T.gemm(A_shared, B_shared, C_local, transpose_B=True)
|
|
|
|
T.copy(C_local, C[by * block_M, bx * block_N])
|
|
|
|
return main
|
|
|
|
|
|
@tilelang.testing.requires_cuda
|
|
@tilelang.testing.requires_cuda_compute_version_eq(8, 0)
|
|
def test_compile_int4_gemm_tgemm():
|
|
func = matmul_nt_int4(1024, 1024, 1024, 128, 128, 64)
|
|
kernel = tilelang.compile(func, out_idx=-1)
|
|
src = kernel.get_kernel_source()
|
|
assert src is not None
|
|
assert "s4.s4.s32" in src or "int4" in src
|
|
|
|
|
|
if __name__ == "__main__":
|
|
tilelang.testing.main()
|