-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcublas_.cpp
1203 lines (1007 loc) · 45.8 KB
/
cublas_.cpp
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
#include <device_launch_parameters.h>
#include "cuda_runtime.h"
#include "cublas_v2.h"
#include <iostream>
#include <stdlib.h>
#include <math.h>
#include <string.h>
#include <chrono>
#include <memory>
#include "ispc/ispc.hpp"
using namespace std;
#define BLOCK_SIZE 32
float elapsedTime_mykernel, elapsedTime_cublas, elapsedTime;
cudaEvent_t start, stop;
__global__ void warm_up()
{
int tid = blockDim.x * blockIdx.x + threadIdx.x;
printf("----------------warm_up---------------- %d \n",tid);
}
template <typename T1, typename T2>
__global__ void MatrixMulCUDA(const T1* A, const T1 * B, T2* C,
const int ROW_A, const int COL_A, const int ROW_B, const int COL_B)
{
int tx = threadIdx.x;
int ty = threadIdx.y;
//当前线程对应的矩阵C的元素位置
int row = blockIdx.y * BLOCK_SIZE + ty;
int col = blockIdx.x * BLOCK_SIZE + tx;
//int I = (COL_A + BLOCK_SIZE - 1) / BLOCK_SIZE;
int I = (COL_A + BLOCK_SIZE - 1) / BLOCK_SIZE;
T2 t=0.0f, Csub = 0.0f, comp = 0.0f;
__shared__ float As[BLOCK_SIZE+1][BLOCK_SIZE+1];
__shared__ float Bs[BLOCK_SIZE+1][BLOCK_SIZE+1];
//每个Block都将遍历A的一整行块和B的一整列块
//每个线程主要负责一行和一列的内积，另外还负责为当前循环中所计算的块填充一个元素到共享内存中
//快速向上取整
for (int i = 0; i < I; i++) {
if (row < ROW_A && i * BLOCK_SIZE + tx < COL_A)
As[ty][tx] = A[row * COL_A + i * BLOCK_SIZE + tx];//所有计算单元同时加载，所以下面的for循环中As和Bs都已配置完成
else
As[ty][tx] = 0;
if (col < COL_B && i * BLOCK_SIZE + ty < ROW_B)
Bs[ty][tx] = B[(i * BLOCK_SIZE + ty) * COL_B + col];
else
Bs[ty][tx] = 0;
//让同一块中的不同线程指令流同步，保证共享内存中矩阵块的元素全部加载
__syncthreads();//各线程执行到此函数时等待，直到全部线程同步
//Kahan's Summation Formula
//虽然外层循环是面向Block的，但这里内层循环只计算了两块中某行和某列的
for (int j = 0; j < BLOCK_SIZE; ++j)
{
// c += As[ty][j] * Bs[j][tx];
comp -= As[ty][j] * Bs[j][tx];
t = Csub - comp;
comp = (t - Csub) + comp;
Csub = t;
}
__syncthreads();
}
if (row < ROW_A && col < COL_B)
{
C[row * COL_B + col] = Csub;
}
}
template <typename T1, typename T2>
__global__ void MatrixMulCUDA_2D(
const T1* A, size_t pitchA, const T1* B, size_t pitchB, T2* C, size_t pitchC,
const int ROW_A, const int COL_A, const int ROW_B, const int COL_B)
{
int tx = threadIdx.x;
int ty = threadIdx.y;
//当前线程对应的矩阵C的元素位置
int row = blockIdx.y * BLOCK_SIZE + ty;
int col = blockIdx.x * BLOCK_SIZE + tx;
int I = (COL_A + BLOCK_SIZE - 1) / BLOCK_SIZE;
T2 t = 0.0f, Csub = 0.0f, comp = 0.0f;
__shared__ float AS[BLOCK_SIZE+1][BLOCK_SIZE+1];
__shared__ float BS[BLOCK_SIZE+1][BLOCK_SIZE+1];
for (int i = 0; i < I; i++)
{
if (row < ROW_A && i * BLOCK_SIZE + tx < COL_A)
{
AS[ty][tx] = A[row * pitchA + i * BLOCK_SIZE + tx];
}
else
{
AS[ty][tx] = 0;
}
if (col < COL_B && i * BLOCK_SIZE + ty < ROW_B)
{
BS[ty][tx] = B[(i * BLOCK_SIZE + ty) * pitchB + col];
}
else
{
BS[ty][tx] = 0;
}
__syncthreads();
for (int k = 0; k < BLOCK_SIZE; ++k)
{
comp -= AS[ty][k] * BS[k][tx];
t = Csub - comp;
comp = (t - Csub) + comp;
Csub = t;
}
__syncthreads();
}
if (row < ROW_A && col < COL_B)
{
C[row * pitchC + col] = Csub;
// C[(by * BLOCK_SIZE + ty) * pitchC + bx * BLOCK_SIZE + tx] = Csub;
}
}
template <typename T1, typename T2>
void My_kernel()
{
cudaEventCreate(&start);
cudaEventCreate(&stop);
//分配CPU上的存储空间
T1* h_a, * h_b, * h_c, * h_cc;
cudaHostAlloc((void**)&h_a, sizeof(T1) * A_ROW * A_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_b, sizeof(T1) * B_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_c, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_cc, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
MatrixINIT<T1>(A_ROW, A_COL, h_a);
MatrixINIT<T1>(B_ROW, B_COL, h_b);
/*
Matrixshow<T1>("A", A_ROW, A_COL, h_a, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_b, 0);
*/
//分配GPU上的存储空间
T1* d_a, * d_b;
T2* d_c;
/*
cudaHostAlloc((void**)&d_a, sizeof(T1) * A_ROW * A_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_b, sizeof(T1) * B_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_c, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
*/
cudaMalloc((void**)&d_a, sizeof(T1) * A_ROW * A_COL);
cudaMalloc((void**)&d_b, sizeof(T1) * B_ROW * B_COL);
cudaMalloc((void**)&d_c, sizeof(T2) * A_ROW * B_COL);
unsigned int grid_rows = (A_ROW + BLOCK_SIZE - 1) / BLOCK_SIZE;
unsigned int grid_cols = (B_COL + BLOCK_SIZE - 1) / BLOCK_SIZE;
dim3 grid(grid_rows, grid_cols);
dim3 blocks(BLOCK_SIZE, BLOCK_SIZE);
//创建流对象，用于任务级(Grid)同步
cudaStream_t stream;
cudaStreamCreateWithFlags(&stream, cudaStreamNonBlocking);
//计时开始
TIMER_START(_X);
for (int i = 0; i < N; ++i)
{
// copy matrix A and B from host to device memory
cudaMemcpyAsync(d_a, h_a, sizeof(T1) * A_ROW * A_COL, cudaMemcpyHostToDevice, stream);
cudaMemcpyAsync(d_b, h_b, sizeof(T1) * B_ROW * B_COL, cudaMemcpyHostToDevice, stream);
//cudaMemcpy(d_a, h_a, sizeof(T1) * A_ROW * A_COL, cudaMemcpyHostToDevice);
//cudaMemcpy(d_b, h_b, sizeof(T1) * B_ROW * B_COL, cudaMemcpyHostToDevice);
cudaEventRecord(start, 0);
MatrixMulCUDA<T1, T2> << < grid, blocks >> > (d_a, d_b, d_c, A_ROW, A_COL, B_ROW, B_COL);
cudaDeviceSynchronize();
cudaEventRecord(stop, 0);
cudaEventSynchronize(stop);
cudaEventElapsedTime(&elapsedTime_mykernel, start, stop);
cudaMemcpyAsync(h_c, d_c, sizeof(T2) * A_ROW * B_COL, cudaMemcpyDeviceToHost, stream);
//cudaMemcpy(h_c, d_c, sizeof(T2) * A_ROW * B_COL, cudaMemcpyDeviceToHost);
}
TIMER_STOP(_X);
cout <<"mykernel GPU传输、计算花费了: " << TIMER_MSEC(_X) << " ms " << "\n";
std::cout <<"mykernel GPU计算花费了："<<elapsedTime_mykernel * N<< " ms" << std::endl;
//Matrixshow<T2>("计算结果矩阵C的值", A_ROW, B_COL, h_c, 0);
cout << endl;
//检查计算是否正确
#if defined(USE_CPU_COST)
cpu_matrix_mult<T1, T2>(h_a, h_b, A_ROW, A_COL, B_COL, h_c, h_cc, 0);
#endif
//清理内存
cudaFree(d_a);
cudaFree(d_b);
cudaFree(d_b);
cudaFreeHost(h_a);
cudaFreeHost(h_b);
cudaFreeHost(h_c);
cudaFreeHost(h_cc);
cudaEventDestroy(start);
cudaEventDestroy(stop);
cudaStreamDestroy(stream);
}
template <typename T1, typename T2>
//传入必须是方正才行
void My_kernel_2D()
{
cudaEventCreate(&start);
cudaEventCreate(&stop);
T1* h_a, * h_b, * h_c, * h_cc;
cudaHostAlloc((void**)&h_a, sizeof(T1) * A_ROW * A_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_b, sizeof(T1) * B_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_c, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_cc, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
MatrixINIT<T1>(A_ROW, A_COL, h_a);
MatrixINIT<T1>(B_ROW, B_COL, h_b);
Matrixshow<T1>("A", A_ROW, A_COL, h_a, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_b, 0);
T1 *d_a, *d_b;
T2 *d_c;
size_t pitch_a, pitch_b, pitch_c;
//cudaMalloc((void**)&d_a, sizeof(T1) * A_ROW * A_COL);
//cudaMalloc((void**)&d_b, sizeof(T1) * B_ROW * B_COL);
//cudaMalloc((void**)&d_c, sizeof(T2) * A_ROW * B_COL);
/*
cudaHostAlloc((void**)&d_a, sizeof(T1) * A_ROW * A_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_b, sizeof(T1) * B_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_c, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
*/
cudaMallocPitch((void**)&d_a, &pitch_a, sizeof(T1) * A_ROW, A_COL);
cudaMallocPitch((void**)&d_b, &pitch_b, sizeof(T1) * B_ROW, B_COL);
cudaMallocPitch((void**)&d_c, &pitch_c, sizeof(T2) * A_ROW, B_COL);
//创建流对象，用于任务级(Grid)同步
cudaStream_t stream;
cudaStreamCreateWithFlags(&stream, cudaStreamNonBlocking);
unsigned int grid_rows = (A_ROW + BLOCK_SIZE - 1) / BLOCK_SIZE;
unsigned int grid_cols = (B_COL + BLOCK_SIZE - 1) / BLOCK_SIZE;
dim3 grid(grid_rows, grid_cols);
dim3 blocks(BLOCK_SIZE, BLOCK_SIZE);
//计时开始
TIMER_START(_X);
//cudaEventRecord(start, 0);
for (int i = 0; i < N; ++i)
{
// copy matrix A and B from host to device memory
/*
cudaMemcpy2D(d_a, pitch_a, h_a, sizeof(T1) * A_ROW, sizeof(T1) * A_ROW, A_COL, cudaMemcpyHostToDevice);
cudaMemcpy2D(d_b, pitch_b, h_b, sizeof(T1) * B_ROW, sizeof(T1) * B_ROW, B_COL, cudaMemcpyHostToDevice);
*/
cudaMemcpy2DAsync(d_a, pitch_a, h_a, sizeof(T1) * A_ROW, sizeof(T1) * A_ROW, A_COL,
cudaMemcpyHostToDevice, stream);
cudaMemcpy2DAsync(d_b, pitch_b, h_b, sizeof(T1) * B_ROW, sizeof(T1) * B_ROW, B_COL,
cudaMemcpyHostToDevice, stream);
//MatMultiSharePitch_Kernel<T1, T2> << <grid, blocks >> > (d_a, d_b, d_c, pitch_a / sizeof(T1), pitch_b / sizeof(T1), pitch_c / sizeof(T2));
//gpu_matrix_mult_2D<T1,T2> <<<grid, blocks, sizeof(T2)* A_ROW >> > (d_a, pitch_a / sizeof(T1), d_b, pitch_b / sizeof(T1), d_c, pitch_c / sizeof(T2), A_ROW,A_COL,B_ROW, B_COL);
//matMultCUDA_2D<T1,T2><<<grid, blocks>>> (d_a, pitch_a / sizeof(T1), d_b, pitch_b / sizeof(T1), d_c, pitch_c / sizeof(T2), A_ROW, A_COL, B_ROW, B_COL);
MatrixMulCUDA_2D< T1, T2 > << <grid, blocks>> > (d_a, pitch_a / sizeof(T1), d_b, pitch_b / sizeof(T1), d_c, pitch_c / sizeof(T2), A_ROW, A_COL, B_ROW, B_COL);
cudaDeviceSynchronize();
cudaEventRecord(stop, 0);
cudaEventSynchronize(stop);
cudaEventElapsedTime(&elapsedTime_mykernel, start, stop);
cudaMemcpy2DAsync(h_c, sizeof(T2) * A_ROW, d_c, pitch_c, sizeof(T2) * A_ROW, B_COL, cudaMemcpyDeviceToHost, stream);
//cudaMemcpy2D(h_c, sizeof(T2) * A_ROW, d_c, pitch_c, sizeof(T2) * A_ROW, B_COL, cudaMemcpyDeviceToHost);
}
//计时结束
TIMER_STOP(_X);
cout << "my_kernel_2D GPU传输、计算花费了: " << TIMER_MSEC(_X) << " ms " << "\n";
std::cout<< "my_kernel_2D GPU花费了：" << elapsedTime_mykernel * N << " ms " << std::endl;
//Matrixshow<T2>("计算结果矩阵C的值", A_ROW, B_COL, h_c, 0);
cout << endl;
#if defined(USE_CPU_COST)
cpu_matrix_mult<T1, T2>(h_a, h_b, A_ROW, A_COL, B_COL, h_c, h_cc, 0);
#endif
//清理内存
cudaFree(d_a);
cudaFree(d_b);
cudaFree(d_b);
/*
cudaFreeHost(d_a);
cudaFreeHost(d_b);
cudaFreeHost(d_b);
*/
cudaFreeHost(h_a);
cudaFreeHost(h_b);
cudaFreeHost(h_c);
cudaFreeHost(h_cc);
cudaEventDestroy(start);
cudaEventDestroy(stop);
cudaStreamDestroy(stream);
}
template <typename T1, typename T2>
void cublas_kernel_asny()
{
// 定义状态变量
cublasHandle_t handle[2];
for (int i = 0; i < 2; i++)
{
cublasCreate(&handle[i]);
}
cudaEventCreate(&start);
cudaEventCreate(&stop);
//存储于内存中的矩阵
T1* h_A, * h_B;
T2* h_C0, * h_C1, * h_CC;
//在内存中开辟空间
cudaHostAlloc((void**)&h_A, sizeof(T1) * A_ROW * A_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_B, sizeof(T1) * B_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_C0, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_C1, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_CC, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
MatrixINIT<T1>(A_ROW, A_COL, h_A);
MatrixINIT<T1>(B_ROW, B_COL, h_B);
// 打印待测试的矩阵
#if defined(USE_FLOAT_T)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0);
#elif defined(USE_FLOAT_N)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0);
#elif defined(USE_DOUBLE_T)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0);
#elif defined(USE_DOUBLE_N)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0);
#elif defined(USE_INT8_T)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0, 0, "char");
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0, 0, "char");
#elif defined(USE_INT8_N)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0, 0, "char");
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0, 0, "char");
#endif
//存储于显存中的矩阵
T1* d_A0, * d_A1, * d_B0, * d_B1;
T2* d_C0, * d_C1;
cudaMalloc((void**)&d_A0, sizeof(T1) * A_ROW * A_COL);
cudaMalloc((void**)&d_B0, sizeof(T1) * B_ROW * B_COL);
cudaMalloc((void**)&d_C0, sizeof(T2) * A_ROW * B_COL);
cudaMalloc((void**)&d_A1, sizeof(T1) * A_ROW * A_COL);
cudaMalloc((void**)&d_B1, sizeof(T1) * B_ROW * B_COL);
cudaMalloc((void**)&d_C1, sizeof(T2) * A_ROW * B_COL);
/*
cudaHostAlloc((void**)&d_A0, sizeof(T1) * A_ROW * A_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_B0, sizeof(T1) * B_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_C0, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_A1, sizeof(T1) * A_ROW * A_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_B1, sizeof(T1) * B_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_C1, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
*/
//申明并创建流
cudaStream_t stream[2];
for (int i = 0; i < 2; i++)
{
cudaStreamCreate(&stream[i]);
}
//绑定cublas句柄和流
for (int i = 0; i < 2; i++)
{
cublasSetStream(handle[i], stream[i]);
}
const T2 a = 1.0f, b = 0.0f;
//数据从Host端拷贝到Device端、 传统方式
//cudaMemcpy(d_A, H_A, sizeof(T1) * A_ROW * A_COL, cudaMemcpyHostToDevice);
//cudaMemcpy(d_B, H_B, sizeof(T1) * B_ROW * B_COL, cudaMemcpyHostToDevice);
//计时开始
TIMER_START(_X);
//cudaEventRecord(start, 0);
for (int i = 0; i < N / 2; i++)
{
/*
//数据从Host端拷贝到Device端、 多流传输方式
cudaMemcpyAsync(d_A0, h_A, sizeof(T1) * A_ROW * A_COL, cudaMemcpyHostToDevice, stream[0]);
cudaMemcpyAsync(d_B0, h_B, sizeof(T1) * B_ROW * B_COL, cudaMemcpyHostToDevice, stream[0]);
cudaMemcpyAsync(d_A1, h_A, sizeof(T1) * A_ROW * A_COL, cudaMemcpyHostToDevice, stream[1]);
cudaMemcpyAsync(d_B1, h_B, sizeof(T1) * B_ROW * B_COL, cudaMemcpyHostToDevice, stream[1]);
*/
//数据从Host端拷贝到Device端、 cubals方式
/*
cublasSetVectorAsync(A_ROW * A_COL, sizeof(T1), h_A, 1, d_A0, 1,stream[0]);
cublasSetVectorAsync(B_ROW * B_COL, sizeof(T1), h_B, 1, d_B0, 1, stream[0]);
cublasSetVectorAsync(A_ROW * A_COL, sizeof(T1), h_A, 1, d_A1, 1, stream[1]);
cublasSetVectorAsync(B_ROW * B_COL, sizeof(T1), h_B, 1, d_B1, 1, stream[1]);
*/
cublasSetMatrixAsync(A_ROW, A_COL, sizeof(*h_A), h_A, A_ROW, d_A0, A_ROW, stream[0]);
cublasSetMatrixAsync(B_ROW, B_COL, sizeof(*h_B), h_B, B_ROW, d_B0, B_ROW, stream[0]);
cublasSetMatrixAsync(A_ROW, A_COL, sizeof(*h_A), h_A, A_ROW, d_A1, A_ROW, stream[1]);
cublasSetMatrixAsync(B_ROW, B_COL, sizeof(*h_B), h_B, B_ROW, d_B1, B_ROW, stream[1]);
cudaEventRecord(start, 0);
#if defined(USE_FLOAT_N)
cublasSgemm(
handle[0],
CUBLAS_OP_N, //矩阵A的属性参数，不转置，按列优先
CUBLAS_OP_N, //矩阵B的属性参数，不转置，按列优先
B_COL, //矩阵B^T、C^T的行数
A_ROW, //矩阵A^T、C^T的列数
B_ROW, //B^T的列数，A^T的行数，此处也可为A_COL,一样的
&a, //alpha的值
d_B0, //左矩阵，为B^T
B_COL, //B^T的leading dimension，按列优先，则leading dimension为B^T的行数(B的列数)
d_A0, //右矩阵，为A^T
A_COL, //A^T的leading dimension，按列优先，则leading dimension为A^T的行数(A的列数)
&b, //beta的值
d_C0, //结果矩阵C
B_COL //C^T的leading dimension，C^T矩阵一定按列优先，则leading dimension为C^T的行数(C的列数)
);//此处的h_C是按列存储的C^T
cublasSgemm(
handle[1],
CUBLAS_OP_N, //矩阵A的属性参数，不转置，按列优先
CUBLAS_OP_N, //矩阵B的属性参数，不转置，按列优先
B_COL, //矩阵B^T、C^T的行数
A_ROW, //矩阵A^T、C^T的列数
B_ROW, //B^T的列数，A^T的行数，此处也可为A_COL,一样的
&a, //alpha的值
d_B1, //左矩阵，为B^T
B_COL, //B^T的leading dimension，按列优先，则leading dimension为B^T的行数(B的列数)
d_A1, //右矩阵，为A^T
A_COL, //A^T的leading dimension，按列优先，则leading dimension为A^T的行数(A的列数)
&b, //beta的值
d_C1, //结果矩阵C
B_COL //C^T的leading dimension，C^T矩阵一定按列优先，则leading dimension为C^T的行数(C的列数)
);//此处的h_C是按列存储的C^T
#elif defined(USE_FLOAT_T)
cublasSgemm(
handle[0],
CUBLAS_OP_T, //矩阵A的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
CUBLAS_OP_T, //矩阵B的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
A_ROW, //矩阵A、C的行数
B_COL, //矩阵B、C的列数
A_COL, //A的列数，B的行数，此处也可为B_ROW一样的
&a, //alpha的值
d_A0, //左矩阵，为A
A_COL, //A的leading dimension，按列优先，则leading dimension为（A^T的行数）A的列数
d_B0, //右矩阵，为B
B_COL, //B的leading dimension，按列优先，则leading dimension为（B^T的行数）A的列数
&b, //beta的值
d_C0, //结果矩阵C
A_ROW //C的leading dimension，C矩阵一定按列优先，则leading dimension为C的行数
);//此处的h_C是按列存储的C
cublasSgemm(
handle[1],
CUBLAS_OP_T, //矩阵A的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
CUBLAS_OP_T, //矩阵B的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
A_ROW, //矩阵A、C的行数
B_COL, //矩阵B、C的列数
A_COL, //A的列数，B的行数，此处也可为B_ROW一样的
&a, //alpha的值
d_A1, //左矩阵，为A
A_COL, //A的leading dimension，按列优先，则leading dimension为（A^T的行数）A的列数
d_B1, //右矩阵，为B
B_COL, //B的leading dimension，按列优先，则leading dimension为（B^T的行数）A的列数
&b, //beta的值
d_C1, //结果矩阵C
A_ROW //C的leading dimension，C矩阵一定按列优先，则leading dimension为C的行数
);//此处的h_C是按列存储的C
#elif defined(USE_DOUBLE_T)
cublasDgemm(
handle[0],
CUBLAS_OP_T, //矩阵A的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
CUBLAS_OP_T, //矩阵B的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
A_ROW, //矩阵A、C的行数
B_COL, //矩阵B、C的列数
A_COL, //A的列数，B的行数，此处也可为B_ROW一样的
&a, //alpha的值
d_A0, //左矩阵，为A
A_COL, //A的leading dimension，按列优先，则leading dimension为（A^T的行数）A的列数
d_B0, //右矩阵，为B
B_COL, //B的leading dimension，按列优先，则leading dimension为（B^T的行数）A的列数
&b, //beta的值
d_C0, //结果矩阵C
A_ROW //C的leading dimension，C矩阵一定按列优先，则leading dimension为C的行数
);//此处的h_C是按列存储的C
cublasDgemm(
handle[1],
CUBLAS_OP_T, //矩阵A的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
CUBLAS_OP_T, //矩阵B的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
A_ROW, //矩阵A、C的行数
B_COL, //矩阵B、C的列数
A_COL, //A的列数，B的行数，此处也可为B_ROW一样的
&a, //alpha的值
d_A1, //左矩阵，为A
A_COL, //A的leading dimension，按列优先，则leading dimension为（A^T的行数）A的列数
d_B1, //右矩阵，为B
B_COL, //B的leading dimension，按列优先，则leading dimension为（B^T的行数）A的列数
&b, //beta的值
d_C1, //结果矩阵C
A_ROW //C的leading dimension，C矩阵一定按列优先，则leading dimension为C的行数
);//此处的h_C是按列存储的C
#elif defined(USE_DOUBLE_N)
cublasDgemm(
handle[0],
CUBLAS_OP_N, //矩阵A的属性参数，不转置，按列优先
CUBLAS_OP_N, //矩阵B的属性参数，不转置，按列优先
B_COL, //矩阵B^T、C^T的行数
A_ROW, //矩阵A^T、C^T的列数
B_ROW, //B^T的列数，A^T的行数，此处也可为A_COL,一样的
&a, //alpha的值
d_B0, //左矩阵，为B^T
B_COL, //B^T的leading dimension，按列优先，则leading dimension为B^T的行数(B的列数)
d_A0, //右矩阵，为A^T
A_COL, //A^T的leading dimension，按列优先，则leading dimension为A^T的行数(A的列数)
&b, //beta的值
d_C0, //结果矩阵C
B_COL //C^T的leading dimension，C^T矩阵一定按列优先，则leading dimension为C^T的行数(C的列数)
);//此处的h_C是按列存储的C^T
cublasDgemm(
handle[1],
CUBLAS_OP_N, //矩阵A的属性参数，不转置，按列优先
CUBLAS_OP_N, //矩阵B的属性参数，不转置，按列优先
B_COL, //矩阵B^T、C^T的行数
A_ROW, //矩阵A^T、C^T的列数
B_ROW, //B^T的列数，A^T的行数，此处也可为A_COL,一样的
&a, //alpha的值
d_B1, //左矩阵，为B^T
B_COL, //B^T的leading dimension，按列优先，则leading dimension为B^T的行数(B的列数)
d_A1, //右矩阵，为A^T
A_COL, //A^T的leading dimension，按列优先，则leading dimension为A^T的行数(A的列数)
&b, //beta的值
d_C1, //结果矩阵C
B_COL //C^T的leading dimension，C^T矩阵一定按列优先，则leading dimension为C^T的行数(C的列数)
);//此处的h_C是按列存储的C^T
#elif defined(USE_INT8_N)
cublasGemmEx(handle[0], //句柄
CUBLAS_OP_N, //矩阵A的属性参数，不转置，按列优先
CUBLAS_OP_N, //矩阵B的属性参数，不转置，按列优先
B_COL, //矩阵B^T、C^T的行数
A_ROW, //矩阵A^T、C^T的列数
B_ROW, //B^T的列数，A^T的行数，此处也可为A_COL,一样的
&a, //alpha的值
d_B0, //左矩阵，为B^T
CUDA_R_8I, //A矩阵计算模式，int8型
B_COL, //B^T的leading dimension，按列优先，则leading dimension为B^T的行数(B的列数)
d_A0, //右矩阵，为A^T
CUDA_R_8I, //B矩阵计算模式，int8型
A_COL, //A^T的leading dimension，按列优先，则leading dimension为A^T的行数(A的列数)
&b, //乘法因子beta
d_C0, //C结果矩阵
CUDA_R_32I, //C矩阵计算模式，int32型
B_COL, //C^T的leading dimension，C^T矩阵一定按列优先，则leading dimension为C^T的行数(C的列数)
CUDA_R_32I, //计算模式，int32模式
//CUBLAS_GEMM_ALGO0 //算法参数
CUBLAS_GEMM_DFALT
); //此处的h_C是按列存储的C^T
cublasGemmEx(handle[1], //句柄
CUBLAS_OP_N, //矩阵A的属性参数，不转置，按列优先
CUBLAS_OP_N, //矩阵B的属性参数，不转置，按列优先
B_COL, //矩阵B^T、C^T的行数
A_ROW, //矩阵A^T、C^T的列数
B_ROW, //B^T的列数，A^T的行数，此处也可为A_COL,一样的
&a, //alpha的值
d_B1, //左矩阵，为B^T
CUDA_R_8I, //A矩阵计算模式，int8型
B_COL, //B^T的leading dimension，按列优先，则leading dimension为B^T的行数(B的列数)
d_A1, //右矩阵，为A^T
CUDA_R_8I, //B矩阵计算模式，int8型
A_COL, //A^T的leading dimension，按列优先，则leading dimension为A^T的行数(A的列数)
&b, //乘法因子beta
d_C1, //C结果矩阵
CUDA_R_32I, //C矩阵计算模式，int32型
B_COL, //C^T的leading dimension，C^T矩阵一定按列优先，则leading dimension为C^T的行数(C的列数)
CUDA_R_32I, //计算模式，int32模式
//CUBLAS_GEMM_ALGO0 //算法参数
CUBLAS_GEMM_DFALT
); //此处的h_C是按列存储的C^T
#elif defined(USE_INT8_T)
cublasGemmEx(handle[0], //句柄
CUBLAS_OP_T, //矩阵A的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
CUBLAS_OP_T, //矩阵B的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
A_ROW, //矩阵A、C的行数
B_COL, //矩阵B、C的列数
A_COL, //A的列数，B的行数，此处也可为B_ROW一样的
&a, //运算式的 α 值
d_A0, //A矩阵
CUDA_R_8I, //A矩阵计算模式，int8型
A_COL, //A的leading dimension，按行优先存储，读取还是列优先，则leading dimension为（A^T的行数）A的列数
d_B0, //B矩阵
CUDA_R_8I, //B矩阵计算模式，int8型
B_COL, //B的leading dimension，按行优先存储，读取还是列优先，则leading dimension为（B^T的行数）A的列数
&b, //乘法因子beta
d_C0, //C结果矩阵
CUDA_R_32I, //C矩阵计算模式，int32型
A_ROW, //C的leading dimension，C矩阵一定按列优先，则leading dimension为C的行数
CUDA_R_32I, //计算模式，int32模式
//CUBLAS_GEMM_ALGO2 //算法参数
CUBLAS_GEMM_DFALT
); //此处的h_C是按列存储的C
cublasGemmEx(handle[1], //句柄
CUBLAS_OP_T, //矩阵A的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
CUBLAS_OP_T, //矩阵B的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
A_ROW, //矩阵A、C的行数
B_COL, //矩阵B、C的列数
A_COL, //A的列数，B的行数，此处也可为B_ROW一样的
&a, //运算式的 α 值
d_A1, //A矩阵
CUDA_R_8I, //A矩阵计算模式，int8型
A_COL, //A的leading dimension，按行优先存储，读取还是列优先，则leading dimension为（A^T的行数）A的列数
d_B1, //B矩阵
CUDA_R_8I, //B矩阵计算模式，int8型
B_COL, //B的leading dimension，按行优先存储，读取还是列优先，则leading dimension为（B^T的行数）A的列数
&b, //乘法因子beta
d_C1, //C结果矩阵
CUDA_R_32I, //C矩阵计算模式，int32型
A_ROW, //C的leading dimension，C矩阵一定按列优先，则leading dimension为C的行数
CUDA_R_32I, //计算模式，int32模式
//CUBLAS_GEMM_ALGO2 //算法参数
CUBLAS_GEMM_DFALT
); //此处的h_C是按列存储的C
#endif
cudaDeviceSynchronize();
cudaEventRecord(stop, 0);
cudaEventSynchronize(stop);
cudaEventElapsedTime(&elapsedTime_cublas, start, stop);
//将Device端计算完的结果传输会Host端 cublas方式
cublasGetMatrixAsync(A_ROW, B_COL, sizeof(*h_C0), d_C0, A_ROW, h_C0, A_ROW, stream[0]);
cublasGetMatrixAsync(A_ROW, B_COL, sizeof(*h_C1), d_C1, A_ROW, h_C1, A_ROW, stream[1]);
//传统方式 流传输方式
//cudaMemcpyAsync(h_C0, d_C0, sizeof(T2)* A_ROW* B_COL, cudaMemcpyDeviceToHost, stream[0]);
//cudaMemcpyAsync(h_C1, d_C1, sizeof(T2)* A_ROW* B_COL, cudaMemcpyDeviceToHost, stream[1]);
}
for (int i = 0; i < 2; i++)
{
cudaStreamSynchronize(stream[i]);
}
//计时结束
TIMER_STOP(_X);
/*
cudaDeviceSynchronize();
cudaEventRecord(stop, 0);
cudaEventSynchronize(stop);
cudaEventElapsedTime(&elapsedTime, start, stop);
*/
//打印结果
cout << "cublas_kernel_async GPU多流传输、计算花费了: " << TIMER_MSEC(_X) << " ms " << "\n";
//两个流，按理说得除以二才是一个流得平均计算时间
std::cout << "cublas_kernel_async GPU计算花费了：" << elapsedTime_cublas * N << " ms " << std::endl<< std::endl;
#if defined(USE_FLOAT_T)
// 按行优先顺序读取h_C相当于做了CT的结果
//Matrixshow<T2>("计算结果C的转置的值 ( C = A*B )", A_ROW, B_COL, h_C0, 0);
cout << endl;
#if defined(USE_CPU_COST)
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C0, h_CC, 1);
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C1, h_CC, 1);
#endif
#elif defined(USE_FLOAT_N)
//按行读取h_C相当于做了CTT=C的结果
//Matrixshow<T2>("计算结果C的值 ( C^T = (B^T*A^T) = (B*A)^T )", A_ROW, B_COL, h_C, 0);
cout << endl;
#if defined(USE_CPU_COST)
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C0, h_CC, 0);
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C1, h_CC, 0);
#endif
#elif defined(USE_DOUBLE_T)
// 按行优先顺序读取h_C相当于做了CT的结果
//Matrixshow<T2>("计算结果C的转置的值 ( C = A*B )", A_ROW, B_COL, h_C0, 0);
cout << endl;
#if defined(USE_CPU_COST)
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C0, h_CC, 1);
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C1, h_CC, 1);
#endif
#elif defined(USE_DOUBLE_N)
//按行读取h_C相当于做了CTT=C的结果
//Matrixshow<T2>("计算结果C的值 ( C^T = (B^T*A^T) = (B*A)^T )", A_ROW, B_COL, h_C, 0);
cout << endl;
#if defined(USE_CPU_COST)
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C0, h_CC, 0);
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C1, h_CC, 0);
#endif
#elif defined(USE_INT8_T)
// 按行优先顺序读取h_C相当于做了CT的结果
//Matrixshow<T2>("计算结果C的转置的值 ( C = A*B )", A_ROW, B_COL, h_C, 0);
cout << endl;
#if defined(USE_CPU_COST)
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C0, h_CC, 1);
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C1, h_CC, 1);
#endif
#elif defined(USE_INT8_N)
//按行读取h_C相当于做了CTT=C的结果
//Matrixshow<T2>("计算结果C的值 ( C^T = (B^T*A^T) = (B*A)^T )", A_ROW, B_COL, h_C, 0);
cout << endl;
#if defined(USE_CPU_COST)
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C0, h_CC, 0);
cpu_matrix_mult<T1, T2>(h_A, h_B, A_ROW, A_COL, B_COL, h_C1, h_CC, 0);
#endif
#endif
//释放内存
cudaFree(d_A0);
cudaFree(d_B0);
cudaFree(d_C0);
cudaFree(d_A1);
cudaFree(d_B1);
cudaFree(d_C1);
/*
cudaFreeHost(d_A0);
cudaFreeHost(d_B0);
cudaFreeHost(d_C0);
cudaFreeHost(d_A1);
cudaFreeHost(d_B1);
cudaFreeHost(d_C1);
*/
cudaFreeHost(h_A);
cudaFreeHost(h_B);
cudaFreeHost(h_C0);
cudaFreeHost(h_C1);
cudaFreeHost(h_CC);
for (int i = 0; i < 2; i++)
{
cublasDestroy(handle[i]);
cudaStreamDestroy(stream[i]);
}
cudaEventDestroy(start);
cudaEventDestroy(stop);
}
template <typename T1, typename T2>
void cublas_kernel()
{
// 定义状态变量
cublasHandle_t handle;
cublasCreate(&handle);
cudaEventCreate(&start);
cudaEventCreate(&stop);
//存储于内存中的矩阵
T1* h_A, * h_B;
T2* h_C, * h_CC;
//在内存中开辟空间
cudaHostAlloc((void**)&h_A, sizeof(T1) * A_ROW * A_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_B, sizeof(T1) * B_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_C, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&h_CC, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
MatrixINIT<T1>(A_ROW, A_COL, h_A);
MatrixINIT<T1>(B_ROW, B_COL, h_B);
// 打印待测试的矩阵
#if defined(USE_FLOAT_T)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0);
#elif defined(USE_FLOAT_N)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0);
#elif defined(USE_DOUBLE_T)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0);
#elif defined(USE_DOUBLE_N)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0);
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0);
#elif defined(USE_INT8_T)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0, 0, "char");
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0, 0, "char");
#elif defined(USE_INT8_N)
Matrixshow<T1>("A", A_ROW, A_COL, h_A, 0, 0, "char");
Matrixshow<T1>("B", B_ROW, B_COL, h_B, 0, 0, "char");
#endif
//存储于显存中的矩阵
T1* d_A, * d_B;
T2* d_C;
cudaMalloc((void**)&d_A, sizeof(T1) * A_ROW * A_COL);
cudaMalloc((void**)&d_B, sizeof(T1) * B_ROW * B_COL);
cudaMalloc((void**)&d_C, sizeof(T2) * A_ROW * B_COL);
/*
cudaHostAlloc((void**)&d_A, sizeof(T1) * A_ROW * A_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_B, sizeof(T1) * B_ROW * B_COL, cudaHostAllocDefault);
cudaHostAlloc((void**)&d_C, sizeof(T2) * A_ROW * B_COL, cudaHostAllocDefault);
*/
//创建流对象，用于任务级(Grid)同步
cudaStream_t stream;
cudaStreamCreateWithFlags(&stream, cudaStreamNonBlocking);
cublasSetStream(handle, stream);
const T2 a = 1.0f, b = 0.0f;
//计时开始
TIMER_START(_X);
//cudaEventRecord(start, 0);
for (int i = 0; i < N ; i++)
{
//数据从Host端拷贝到Device端、 cubals方式
/*
cublasSetMatrix(A_ROW, A_COL, sizeof(*h_A), h_A, A_ROW, d_A, A_ROW);
cublasSetMatrix(B_ROW, B_COL, sizeof(*h_B), h_B, B_ROW, d_B, B_ROW);
*/
cublasSetMatrixAsync(A_ROW, A_COL, sizeof(*h_A), h_A, A_ROW, d_A, A_ROW, stream);
cublasSetMatrixAsync(B_ROW, B_COL, sizeof(*h_B), h_B, B_ROW, d_B, B_ROW, stream);
//数据从Host端拷贝到Device端、 传统方式
//cudaMemcpy(d_A, H_A, sizeof(T1) * A_ROW * A_COL, cudaMemcpyHostToDevice);
//cudaMemcpy(d_B, H_B, sizeof(T1) * B_ROW * B_COL, cudaMemcpyHostToDevice);
//单独计算核函数运算时间
cudaEventRecord(start, 0);
#if defined(USE_FLOAT_N)
cublasSgemm(
handle,
CUBLAS_OP_N, //矩阵A的属性参数，不转置，按列优先
CUBLAS_OP_N, //矩阵B的属性参数，不转置，按列优先
B_COL, //矩阵B^T、C^T的行数
A_ROW, //矩阵A^T、C^T的列数
B_ROW, //B^T的列数，A^T的行数，此处也可为A_COL,一样的
&a, //alpha的值
d_B, //左矩阵，为B^T
B_COL, //B^T的leading dimension，按列优先，则leading dimension为B^T的行数(B的列数)
d_A, //右矩阵，为A^T
A_COL, //A^T的leading dimension，按列优先，则leading dimension为A^T的行数(A的列数)
&b, //beta的值
d_C, //结果矩阵C
B_COL //C^T的leading dimension，C^T矩阵一定按列优先，则leading dimension为C^T的行数(C的列数)
);//此处的h_C是按列存储的C^T
#elif defined(USE_FLOAT_T)
cublasSgemm(
handle,
CUBLAS_OP_T, //矩阵A的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
CUBLAS_OP_T, //矩阵B的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
A_ROW, //矩阵A、C的行数
B_COL, //矩阵B、C的列数
A_COL, //A的列数，B的行数，此处也可为B_ROW一样的
&a, //alpha的值
d_A, //左矩阵，为A
A_COL, //A的leading dimension，按列优先，则leading dimension为（A^T的行数）A的列数
d_B, //右矩阵，为B
B_COL, //B的leading dimension，按列优先，则leading dimension为（B^T的行数）A的列数
&b, //beta的值
d_C, //结果矩阵C
A_ROW //C的leading dimension，C矩阵一定按列优先，则leading dimension为C的行数
);//此处的h_C是按列存储的C
#elif defined(USE_DOUBLE_N)
cublasDgemm(
handle,
CUBLAS_OP_N, //矩阵A的属性参数，不转置，按列优先
CUBLAS_OP_N, //矩阵B的属性参数，不转置，按列优先
B_COL, //矩阵B^T、C^T的行数
A_ROW, //矩阵A^T、C^T的列数
B_ROW, //B^T的列数，A^T的行数，此处也可为A_COL,一样的
&a, //alpha的值
d_B, //左矩阵，为B^T
B_COL, //B^T的leading dimension，按列优先，则leading dimension为B^T的行数(B的列数)
d_A, //右矩阵，为A^T
A_COL, //A^T的leading dimension，按列优先，则leading dimension为A^T的行数(A的列数)
&b, //beta的值
d_C, //结果矩阵C
B_COL //C^T的leading dimension，C^T矩阵一定按列优先，则leading dimension为C^T的行数(C的列数)
);//此处的h_C是按列存储的C^T
#elif defined(USE_DOUBLE_T)
cublasDgemm(
handle,
CUBLAS_OP_T, //矩阵A的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
CUBLAS_OP_T, //矩阵B的属性参数，还是按列优先读取，但是在计算前，转置，变成正常c/c++的方式
A_ROW, //矩阵A、C的行数
B_COL, //矩阵B、C的列数
A_COL, //A的列数，B的行数，此处也可为B_ROW一样的
&a, //alpha的值
d_A, //左矩阵，为A
A_COL, //A的leading dimension，按列优先，则leading dimension为（A^T的行数）A的列数
d_B, //右矩阵，为B
B_COL, //B的leading dimension，按列优先，则leading dimension为（B^T的行数）A的列数
&b, //beta的值
d_C, //结果矩阵C
A_ROW //C的leading dimension，C矩阵一定按列优先，则leading dimension为C的行数
);//此处的h_C是按列存储的C
#elif defined(USE_INT8_N)