biondizzle/nvfp4-megamoe-kernel
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

WIP: TMEM vector for per-row row_sum (not yet working)

Browse Source 
Key finding: the root cause is that each epilogue thread owns MULTIPLE rows
in the QK C-fragment, so scalar row_max/row_sum are wrong (global across
all rows, not per-row). The V=ones diagnostic confirmed: all 128 threads
use the same row_sum (from row 114).

Tried: TMEM vector store+load of row_sum (composition(tStS, (128,2))).
This is a no-op because both write and read use the SAME QK partition
with a scalar row_sum. The vector approach only helps when different
partitions are used for write vs read, or when per-row values are stored.

Next steps:
1. Need PER-ROW row_max and row_sum, not per-thread scalar
2. The CUTLASS FMHA works because each thread owns exactly 1 row
3. Options: restructure thread layout, or compute per-row values differently
4. The vector must store ALL 128 per-row values, then read per-row in C9
This commit is contained in:
biondizzle
2026-05-21 18:45:30 +00:00
parent 331d9e95f3
commit a983a8fb41
1 changed files with 62 additions and 28 deletions
Download Patch File Download Diff File Expand all files Collapse all files

90
tests/unit/test_fmha_v3_softmax.py
View File

@@ -53,6 +53,7 @@ class FmhaV3Softmax:
s_cols = self.qk_mma_tiler[1] # 128
o_after = max(s_cols, p_end) # 128
self.tmem_o0_offset = ((o_after + 31) // 32) * 32 # align to 32 = 128
self.tmem_vec_offset = 0 # Reuse S region (free after softmax loop)
o_cols = find_tmem_tensor_col_offset(tOtO) # footprint of O
total = self.tmem_o0_offset + o_cols
# Must be multiple of 32 AND power of 2
@@ -70,8 +71,8 @@ class FmhaV3Softmax:
self.q_dtype = q.element_type; self.o_dtype = c.element_type; self.c_dtype = self.o_dtype
self.a_major = LayoutEnum.from_tensor(q).mma_major_mode()
self.b_major = LayoutEnum.from_tensor(k).mma_major_mode()
# s_k = cute.size(v, mode=[0])
# FMHA-style V: reconstruct as (HEAD_DIM, 128, 1) MN-major
# s_k = cute.size(v, mode=[0]) # BROKEN in @cute.jit
# FMHA-style V: reconstruct as (HEAD_DIM, s_k, 1) MN-major
v_fmha = cute.make_tensor(
v.iterator,
cute.make_layout(
@@ -232,6 +233,26 @@ class FmhaV3Softmax:
tScP = cute.make_tensor(tScS.iterator, tScP_layout)
tTMEM_STOREcP = thr_store.partition_S(tScP)
# --- Vector TMEM (per-row row_sum storage, FMHA pattern) ---
# composition(tStS.layout, (128, 2)) = 2 FP32 columns per logical row
# vec[0] = row_sum (final, after loop), vec[1] = unused
# Reuses S TMEM region (offset 0), free after softmax loop writes
tStS_vec_layout = cute.composition(tStS.layout, cute.make_layout((128, 2)))
tStS_vec = cute.make_tensor(tStS.iterator + self.tmem_vec_offset, tStS_vec_layout)
tScS_vec_layout = cute.composition(tScS.layout, cute.make_layout((128, 2)))
tScS_vec = cute.make_tensor(tScS.iterator, tScS_vec_layout)
tmem_store_vec_atom = cute.make_copy_atom(tcgen05.copy.St32x32bOp(tcgen05.copy.Repetition(2)), self.qk_acc_dtype)
tiled_tmem_store_vec = tcgen05.make_tmem_copy(tmem_store_vec_atom, tStS_vec)
thr_tmem_store_vec = tiled_tmem_store_vec.get_slice(sfw_idx)
tTMEM_STORE_VECtS = thr_tmem_store_vec.partition_D(tStS_vec)
tTMEM_STORE_VECcS = thr_tmem_store_vec.partition_S(tScS_vec)
tmem_load_vec_atom = cute.make_copy_atom(tcgen05.copy.Ld32x32bOp(tcgen05.copy.Repetition(2)), self.qk_acc_dtype)
tiled_tmem_load_vec = tcgen05.make_tmem_copy(tmem_load_vec_atom, tStS_vec)
thr_tmem_load_vec = tiled_tmem_load_vec.get_slice(sfw_idx)
tTMEM_LOAD_VECtS = thr_tmem_load_vec.partition_S(tStS_vec)
tTMEM_LOAD_VECcS = thr_tmem_load_vec.partition_D(tScS_vec)
# --- C6: O TMEM load/store for rescale (correction_rescale pattern) ---
corr_tile_size = 16
cO = cute.make_identity_tensor((self.pv_mma_tiler[0], self.pv_mma_tiler[1]))
@@ -354,8 +375,19 @@ class FmhaV3Softmax:
# --- C9: Final normalization via O TMEM rescale ---
pv_done_bar.arrive_and_wait()
inv_row_sum = cutlass.Float32(1.0) / row_sum
# Store final row_sum to TMEM vector (per-row, using QK partition)
tTMEM_STORE_VECrS_final = cute.make_rmem_tensor(tTMEM_STORE_VECcS.shape, self.qk_acc_dtype)
tTMEM_STORE_VECrS_final[0] = row_sum
cute.copy(tiled_tmem_store_vec, tTMEM_STORE_VECrS_final, tTMEM_STORE_VECtS)
cute.arch.fence_view_async_tmem_store()
# Read vector back: per-row row_sum using QK partition coordinates
tTMEM_LOAD_VECrS = cute.make_rmem_tensor(tTMEM_LOAD_VECcS.shape, self.qk_acc_dtype)
cute.copy(tiled_tmem_load_vec, tTMEM_LOAD_VECtS, tTMEM_LOAD_VECrS)
cute.arch.fence_view_async_tmem_load()
inv_row_sum = cutlass.Float32(1.0) / tTMEM_LOAD_VECrS[0]
# Normalize O in TMEM
tTMrO_final = cute.make_rmem_tensor((tTMEM_LOADcO.shape, o_col_tiles), self.qk_acc_dtype)
for i in range(o_col_tiles):
tTMrO_i_ = tTMrO_final[None, i]
@@ -384,34 +416,36 @@ class FmhaV3Softmax:
tmem.relinquish_alloc_permit()
tmem.free(tmem_ptr)
def test():
import math
torch.manual_seed(42)
n = 128; m = 128; hd = HEAD_DIM
q = torch.randn(m, hd, 1, dtype=torch.bfloat16, device="cuda")
k = torch.randn(n, hd, 1, dtype=torch.bfloat16, device="cuda")
v = torch.randn(n, hd, dtype=torch.bfloat16, device="cuda")
v_kernel = v.unsqueeze(-1)
c = torch.zeros(m, hd, 1, dtype=torch.bfloat16, device="cuda")
qf = q[:,:,0].float(); kf = k[:,:,0].float()
attn = qf @ kf.T / math.sqrt(hd)
ref = torch.softmax(attn, dim=-1) @ v.float()
mQ = ct.from_dlpack(q).mark_layout_dynamic(leading_dim=ct.get_leading_dim(q))
mK = ct.from_dlpack(k).mark_layout_dynamic(leading_dim=ct.get_leading_dim(k))
mV = ct.from_dlpack(v_kernel).mark_layout_dynamic(leading_dim=ct.get_leading_dim(v_kernel))
mC = ct.from_dlpack(c).mark_layout_dynamic(leading_dim=ct.get_leading_dim(c))
stream = cuda.CUstream(torch.cuda.current_stream().cuda_stream)
kernel = FmhaV3Softmax()
print("Compiling...", flush=True)
compiled = cute.compile(kernel, mQ, mK, mV, mC, stream)
print("Running n=128...", flush=True)
compiled(mQ, mK, mV, mC, stream)
torch.cuda.synchronize()
out = c[:,:,0].float()
cos = torch.nn.functional.cosine_similarity(out.flatten().unsqueeze(0), ref.flatten().unsqueeze(0)).item()
max_err = (out - ref).abs().max().item()
print(f"n=128: cosine {cos:.6f} max_err {max_err:.6f}")
print(f"out[0,:4]={out[0,:4].tolist()} ref[0,:4]={ref[0,:4].tolist()}")
for n in [128, 256, 384]:
m, hd = 128, HEAD_DIM
q = torch.randn(m, hd, 1, dtype=torch.bfloat16, device="cuda")
k = torch.randn(n, hd, 1, dtype=torch.bfloat16, device="cuda")
v = torch.randn(n, hd, dtype=torch.bfloat16, device="cuda")
v_kernel = v.unsqueeze(-1)
c = torch.zeros(m, hd, 1, dtype=torch.bfloat16, device="cuda")
qf = q[:,:,0].float(); kf = k[:,:,0].float()
attn = qf @ kf.T / math.sqrt(hd)
ref = torch.softmax(attn, dim=-1) @ v.float()
mQ = ct.from_dlpack(q).mark_layout_dynamic(leading_dim=ct.get_leading_dim(q))
mK = ct.from_dlpack(k).mark_layout_dynamic(leading_dim=ct.get_leading_dim(k))
mV = ct.from_dlpack(v_kernel).mark_layout_dynamic(leading_dim=ct.get_leading_dim(v_kernel))
mC = ct.from_dlpack(c).mark_layout_dynamic(leading_dim=ct.get_leading_dim(c))
stream = cuda.CUstream(torch.cuda.current_stream().cuda_stream)
kernel = FmhaV3Softmax()
print(f"n={n}: Compiling...", flush=True)
compiled = cute.compile(kernel, mQ, mK, mV, mC, stream)
print(f"n={n}: tmem: s0={kernel.tmem_s0_offset} p0={kernel.tmem_p0_offset} o0={kernel.tmem_o0_offset} vec={kernel.tmem_vec_offset} alloc={kernel.num_tmem_alloc_cols}", flush=True)
print(f"n={n}: Running...", flush=True)
compiled(mQ, mK, mV, mC, stream)
torch.cuda.synchronize()
out = c[:,:,0].float()
cos = torch.nn.functional.cosine_similarity(out.flatten().unsqueeze(0), ref.flatten().unsqueeze(0)).item()
max_err = (out - ref).abs().max().item()
print(f"FMHA softmax n={n}: cosine {cos:.6f} max_err {max_err:.6f} {'PASS' if cos >= 0.999 else 'FAIL'}", flush=True)
if __name__ == "__main__":
test()