ywkang
049e3d8bb3
Move benches/ -> src/kernbench/benches/ and src/kernbench/cli/probe.py -> src/kernbench/probes/probe.py. Each bench self-registers via @bench(name=..., description=...); kernbench list enumerates benches with auto-assigned indices, --bench accepts kebab-case name or numeric index. Audit at package-import time fails if any non-underscore module forgets the decorator. ADR-0010 (EN + KO) updated to reflect the new resolver path, list subcommand, and probes package separation. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
47 lines
1.8 KiB
Python
47 lines
1.8 KiB
Python
"""QKV GEMM benchmark: Q*K^T projection on all PEs in a cube (multi-PE).
|
|
|
|
Column-parallel GEMM: a is replicated (cube-level), b/out are column-sharded.
|
|
M_CPU fans out KernelLaunchMsg to all 8 PE_CPUs (ADR-0009 D3).
|
|
|
|
Kernel: tl.load(a) + tl.ref(b) + tl.composite(gemm) + tl.wait()
|
|
- Tensor a is loaded into TCM via DMA
|
|
- Tensor b stays in HBM; PE_SCHEDULER streams it per-tile (32x64x32)
|
|
"""
|
|
from kernbench.benches.registry import bench
|
|
from kernbench.policy.placement.dp import DPPolicy
|
|
|
|
# GEMM dimensions: (M, K) x (K, N) -> (M, N)
|
|
# Small dims (1 tile) for fast regression. The test verifies the multi-PE
|
|
# fan-out pipeline, not large-matrix throughput.
|
|
M, K, N = 32, 64, 32
|
|
DTYPE = "f16"
|
|
|
|
|
|
def _gemm_kernel(a_ptr, b_ptr, out_ptr, M, K, N, tl, DTYPE="f16"):
|
|
"""QKV GEMM kernel: out = a @ b.
|
|
|
|
a is loaded into TCM (DMA_READ).
|
|
b is referenced in HBM (tl.ref, no DMA -- scheduler streams per-tile).
|
|
"""
|
|
a = tl.load(a_ptr, shape=(M, K), dtype=DTYPE)
|
|
b = tl.ref(b_ptr, shape=(K, N), dtype=DTYPE)
|
|
handle = tl.composite(op="gemm", a=a, b=b, out_ptr=out_ptr)
|
|
tl.wait(handle)
|
|
|
|
|
|
@bench(
|
|
name="qkv-gemm-multi-pe",
|
|
description="Column-parallel QKV GEMM across all PEs in a cube (multi-PE).",
|
|
)
|
|
def run(torch):
|
|
"""Run the multi-PE QKV GEMM benchmark."""
|
|
# DP placement: a=replicate (cube-level), b/out=column_wise (N-axis split)
|
|
a = torch.zeros((M, K), dtype=DTYPE, dp=DPPolicy(cube="replicate", pe="replicate"), name="a")
|
|
b = torch.zeros((K, N), dtype=DTYPE, dp=DPPolicy(cube="replicate", pe="column_wise"), name="b")
|
|
out = torch.empty(
|
|
(M, N), dtype=DTYPE, dp=DPPolicy(cube="replicate", pe="column_wise"), name="out",
|
|
)
|
|
|
|
# Launch GEMM kernel on all PEs
|
|
torch.launch("qkv_gemm_multi", _gemm_kernel, a, b, out, M, K, N)
|