mirror of
https://github.com/tile-ai/tilelang.git
synced 2026-10-28 07:43:01 +08:00
44 lines
1.4 KiB
Python
44 lines
1.4 KiB
Python
import tilelang
|
|
import tilelang.language as T
|
|
import torch
|
|
|
|
|
|
def tilelang_copy(M, N, block_M, block_N, dtype="float16"):
|
|
# add decorator @tilelang.jit if you want to return a torch function
|
|
@T.prim_func
|
|
def main(
|
|
A: T.Tensor((M, N), dtype),
|
|
B: T.Tensor((M, N), dtype),
|
|
):
|
|
# Initialize Kernel Context
|
|
with T.Kernel(T.ceildiv(N, block_N), T.ceildiv(M, block_M), threads=128) as (bx, by):
|
|
for i, j in T.Parallel(block_M, block_N):
|
|
B[by * block_M + i, bx * block_N + j] = A[by * block_M + i, bx * block_N + j]
|
|
|
|
return main
|
|
|
|
|
|
def run_tilelang_copy(M=1024, N=1024, block_M=128, block_N=128, dtype="float16"):
|
|
program = tilelang_copy(M, N, block_M, block_N, dtype)
|
|
kernel = tilelang.compile(
|
|
program,
|
|
out_idx=[1],
|
|
target="cuda",
|
|
pass_configs={
|
|
"tl.disable_warp_specialized": True,
|
|
"tl.disable_tma_lower": True
|
|
})
|
|
a = torch.randn(M, N, device="cuda", dtype=getattr(torch, dtype))
|
|
b = kernel(a)
|
|
torch.testing.assert_close(b, a, rtol=1e-2, atol=1e-2)
|
|
|
|
|
|
def test_tilelang_copy():
|
|
run_tilelang_copy(M=1024, N=1024, block_M=128, block_N=128)
|
|
run_tilelang_copy(M=1024, N=576, block_M=32, block_N=576)
|
|
run_tilelang_copy(M=1024, N=576, block_M=32, block_N=576, dtype="float")
|
|
|
|
|
|
if __name__ == "__main__":
|
|
tilelang.testing.main()
|