[Kernel] Integrate CUTLASS MoE kernel with PPLX (#18762)

Signed-off-by: ElizaWszola <ewszola@redhat.com>
Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>
Co-authored-by: Tyler Michael Smith <tyler@neuralmagic.com>
This commit is contained in:
ElizaWszola
2025-06-06 18:26:11 -07:00
committed by GitHub
co-authored by Tyler Michael Smith
parent 6e0cd10f72
commit 84166fee97
26 changed files with 917 additions and 408 deletions
@@ -632,7 +632,8 @@ def test_cutlass_fp8_group_gemm(num_experts: int, per_act_token: bool,
ops.cutlass_moe_mm(out_tensors_stacked, a_tensors_stacked,
b_tensors_stacked, a_scales_tensors_stacked,
b_scales_tensors_stacked, expert_offsets[:-1],
problem_sizes, ab_strides, ab_strides, c_strides)
problem_sizes, ab_strides, ab_strides, c_strides,
per_act_token, per_out_ch)
# Validate each group's result against the baseline
for g in range(num_experts):