Skip to content

Commit 87ec4ba

Browse files
28Smiles28Smiles
committed
feat: SM_75 Support
1 parent 7901983 commit 87ec4ba

File tree

2 files changed

+75
-3
lines changed

2 files changed

+75
-3
lines changed

‎awq/kernels/csrc/quantization/gemm_cuda_gen.cu

Lines changed: 52 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -200,22 +200,72 @@ __global__ void __launch_bounds__(128) gemm_forward_4bit_cuda_m128n64k32(int spl
200200

201201
for (int i_0_3 = 0; i_0_3 < 4; ++i_0_3) {
202202
for (int j_0_4 = 0; j_0_4 < 2; ++j_0_4) {
203+
#if defined(__CUDA_ARCH__) && __CUDA_ARCH__ < 800
204+
{
205+
__asm__ __volatile__(
206+
"mma.sync.aligned.m16n8k8.row.col.f32.f16.f16.f32"
207+
"{%0, %1, %2, %3}, {%4, %5}, {%6}, {%7, %8, %9, %10};\n"
208+
: "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[0]), "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[1]), "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[2]), "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[3])
209+
: "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[0]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[1]),
210+
"r"(((unsigned *)(B_shared_warp + (j_0_4 * 8)))[0]),
211+
"f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[0]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[1]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[2]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[3])
212+
);
213+
}
214+
215+
{
216+
__asm__ __volatile__(
217+
"mma.sync.aligned.m16n8k8.row.col.f32.f16.f16.f32"
218+
"{%0, %1, %2, %3}, {%4, %5}, {%6}, {%7, %8, %9, %10};\n"
219+
: "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[0]), "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[1]), "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[2]), "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[3])
220+
: "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[2]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[3]),
221+
"r"(((unsigned *)(B_shared_warp + (j_0_4 * 8)))[1]),
222+
"f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[0]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[1]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[2]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[3])
223+
);
224+
}
203225

226+
{
227+
__asm__ __volatile__(
228+
"mma.sync.aligned.m16n8k8.row.col.f32.f16.f16.f32"
229+
"{%0, %1, %2, %3}, {%4, %5}, {%6}, {%7, %8, %9, %10};\n"
230+
: "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[0]), "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[1]), "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[2]), "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[3])
231+
: "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[0]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[1]),
232+
"r"(((unsigned *)(B_shared_warp + ((j_0_4 * 8) + 4)))[0]),
233+
"f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[0]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[1]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[2]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[3])
234+
);
235+
}
236+
{
237+
__asm__ __volatile__(
238+
"mma.sync.aligned.m16n8k8.row.col.f32.f16.f16.f32"
239+
"{%0, %1, %2, %3}, {%4, %5}, {%6}, {%7, %8, %9, %10};\n"
240+
: "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[0]), "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[1]), "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[2]), "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[3])
241+
: "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[2]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[3]),
242+
"r"(((unsigned *)(B_shared_warp + ((j_0_4 * 8) + 4)))[1]),
243+
"f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[0]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[1]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[2]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[3])
244+
);
245+
}
246+
#else
204247
{
205248
__asm__ __volatile__(
206249
"mma.sync.aligned.m16n8k16.row.col.f32.f16.f16.f32"
207250
"{%0, %1, %2, %3}, {%4, %5, %6, %7}, {%8, %9}, {%10, %11, %12, %13};\n"
208251
: "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[0]), "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[1]), "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[2]), "=f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[3])
209-
: "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[0]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[1]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[2]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[3]), "r"(((unsigned *)(B_shared_warp + (j_0_4 * 8)))[0]), "r"(((unsigned *)(B_shared_warp + (j_0_4 * 8)))[1]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[0]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[1]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[2]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[3]));
252+
: "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[0]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[1]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[2]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[3]),
253+
"r"(((unsigned *)(B_shared_warp + (j_0_4 * 8)))[0]), "r"(((unsigned *)(B_shared_warp + (j_0_4 * 8)))[1]),
254+
"f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[0]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[1]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[2]), "f"(((float *)(C_warp + ((i_0_3 * 16) + (j_0_4 * 8))))[3])
255+
);
210256
}
211257

212258
{
213259
__asm__ __volatile__(
214260
"mma.sync.aligned.m16n8k16.row.col.f32.f16.f16.f32"
215261
"{%0, %1, %2, %3}, {%4, %5, %6, %7}, {%8, %9}, {%10, %11, %12, %13};\n"
216262
: "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[0]), "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[1]), "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[2]), "=f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[3])
217-
: "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[0]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[1]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[2]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[3]), "r"(((unsigned *)(B_shared_warp + ((j_0_4 * 8) + 4)))[0]), "r"(((unsigned *)(B_shared_warp + ((j_0_4 * 8) + 4)))[1]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[0]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[1]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[2]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[3]));
263+
: "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[0]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[1]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[2]), "r"(((unsigned *)(A_shared_warp + (i_0_3 * 8)))[3]),
264+
"r"(((unsigned *)(B_shared_warp + ((j_0_4 * 8) + 4)))[0]), "r"(((unsigned *)(B_shared_warp + ((j_0_4 * 8) + 4)))[1]),
265+
"f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[0]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[1]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[2]), "f"(((float *)(C_warp + (((i_0_3 * 16) + (j_0_4 * 8)) + 4)))[3])
266+
);
218267
}
268+
#endif
219269
}
220270
}
221271
}

‎awq/kernels/csrc/quantization_new/gemm/gemm_cuda.cu

Lines changed: 23 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -120,11 +120,33 @@ __inline__ __device__ void cp_async_cg_A(uint32_t smem_int_ptr, const uint4 *__r
120120

121121
__device__ __inline__ void mma_m16n8k16(float *C_warp, half *A_shared_warp, half *B_shared_warp)
122122
{
123+
#if defined(__CUDA_ARCH__) && __CUDA_ARCH__ < 800
124+
__asm__ __volatile__(
125+
"mma.sync.aligned.m16n8k8.row.col.f32.f16.f16.f32"
126+
"{%0, %1, %2, %3}, {%4, %5}, {%6}, {%7, %8, %9, %10};"
127+
: "=f"(((float *)C_warp)[0]), "=f"(((float *)C_warp)[1]), "=f"(((float *)C_warp)[2]), "=f"(((float *)C_warp)[3])
128+
: "r"(((unsigned *)A_shared_warp)[0]), "r"(((unsigned *)A_shared_warp)[1]),
129+
"r"(((unsigned *)B_shared_warp)[0]),
130+
"f"(((float *)C_warp)[0]), "f"(((float *)C_warp)[1]), "f"(((float *)C_warp)[2]), "f"(((float *)C_warp)[3])
131+
);
132+
__asm__ __volatile__(
133+
"mma.sync.aligned.m16n8k8.row.col.f32.f16.f16.f32"
134+
"{%0, %1, %2, %3}, {%4, %5}, {%6}, {%7, %8, %9, %10};"
135+
: "=f"(((float *)C_warp)[0]), "=f"(((float *)C_warp)[1]), "=f"(((float *)C_warp)[2]), "=f"(((float *)C_warp)[3])
136+
: "r"(((unsigned *)A_shared_warp)[2]), "r"(((unsigned *)A_shared_warp)[3]),
137+
"r"(((unsigned *)B_shared_warp)[1]),
138+
"f"(((float *)C_warp)[0]), "f"(((float *)C_warp)[1]), "f"(((float *)C_warp)[2]), "f"(((float *)C_warp)[3])
139+
);
140+
#else
123141
__asm__ __volatile__(
124142
"mma.sync.aligned.m16n8k16.row.col.f32.f16.f16.f32"
125143
"{%0, %1, %2, %3}, {%4, %5, %6, %7}, {%8, %9}, {%10, %11, %12, %13};"
126144
: "=f"(((float *)C_warp)[0]), "=f"(((float *)C_warp)[1]), "=f"(((float *)C_warp)[2]), "=f"(((float *)C_warp)[3])
127-
: "r"(((unsigned *)A_shared_warp)[0]), "r"(((unsigned *)A_shared_warp)[1]), "r"(((unsigned *)A_shared_warp)[2]), "r"(((unsigned *)A_shared_warp)[3]), "r"(((unsigned *)B_shared_warp)[0]), "r"(((unsigned *)B_shared_warp)[1]), "f"(((float *)C_warp)[0]), "f"(((float *)C_warp)[1]), "f"(((float *)C_warp)[2]), "f"(((float *)C_warp)[3]));
145+
: "r"(((unsigned *)A_shared_warp)[0]), "r"(((unsigned *)A_shared_warp)[1]), "r"(((unsigned *)A_shared_warp)[2]), "r"(((unsigned *)A_shared_warp)[3]),
146+
"r"(((unsigned *)B_shared_warp)[0]), "r"(((unsigned *)B_shared_warp)[1]),
147+
"f"(((float *)C_warp)[0]), "f"(((float *)C_warp)[1]), "f"(((float *)C_warp)[2]), "f"(((float *)C_warp)[3])
148+
);
149+
#endif
128150
}
129151

130152
template <int CTA_M, int CTA_N, int CTA_K, int CTA_SIZE, int SHARED_K_ITERS, int STAGES>

0 commit comments

Comments
 (0)