ops_gpu -> ops_cl (#12103)

This commit is contained in:
chenyu
2025-09-10 15:15:48 -04:00
committed by GitHub
parent 0599e86186
commit 0e266f376c
43 changed files with 100 additions and 100 deletions

View File

@@ -93,7 +93,7 @@ class TestKernelOpts(unittest.TestCase):
a = Tensor.rand(8, N, 8, N)
r = a.sum(axis=(1,3))
helper_linearizer_opt(r, [
# openCL / GPU=1 is 256 max threads
# openCL / CL=1 is 256 max threads
[Opt(OptOps.GROUPTOP, 0, 2)], [Opt(OptOps.GROUPTOP, 0, 32)],
[Opt(OptOps.GROUPTOP, 1, 2)], [Opt(OptOps.GROUPTOP, 1, 32)], # Checking how it works with 1 grouped_reduce.
[Opt(OptOps.GROUPTOP, 0, 2), Opt(OptOps.GROUPTOP, 1, 2)],