forked from ekondis/mixbench
/
mix_kernels_hip.cpp
242 lines (207 loc) · 8.76 KB
/
mix_kernels_hip.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
/**
* mix_kernels_hip.cpp: This file is part of the mixbench GPU micro-benchmark suite.
*
* Contact: Elias Konstantinidis <ekondis@gmail.com>
**/
#include "hip/hip_runtime.h"
#include <stdio.h>
#ifdef __CUDACC__
#include <math_constants.h>
#define GPU_INF(_T) (_T)(CUDART_INF)
#else
#include <limits>
#define GPU_INF(_T) std::numeric_limits<_T>::infinity()
#endif
#include "lhiputil.h"
#define COMP_ITERATIONS (8192)
#define UNROLL_ITERATIONS (32)
#define REGBLOCK_SIZE (8)
#define UNROLLED_MEMORY_ACCESSES (UNROLL_ITERATIONS/2)
template <class T>
class functor_mad{
public:
T operator()(T a, T b, T c){
return a * b + c;
}
};
template<>
double functor_mad<double>::operator()(double a, double b, double c){
return fma(a, b, c);
}
template <class T, int blockdim, int memory_ratio>
__global__ void
benchmark_func(hipLaunchParm lp, T seed, volatile T *g_data){
functor_mad<T> mad_op;
const int index_stride = blockdim;
const int index_base = hipBlockIdx_x*blockdim*UNROLLED_MEMORY_ACCESSES + hipThreadIdx_x;
const int halfarraysize = hipGridDim_x*blockdim*UNROLLED_MEMORY_ACCESSES;
const int offset_slips = 1+UNROLLED_MEMORY_ACCESSES-((memory_ratio+1)/2);
const int array_index_bound = index_base+offset_slips*index_stride;
const int initial_index_range = memory_ratio>0 ? UNROLLED_MEMORY_ACCESSES % ((memory_ratio+1)/2) : 1;
int initial_index_factor = 0;
volatile T *data = g_data;
int array_index = index_base;
T r0 = seed + hipBlockIdx_x * blockdim + hipThreadIdx_x,
r1 = r0+(T)(2),
r2 = r0+(T)(3),
r3 = r0+(T)(5),
r4 = r0+(T)(7),
r5 = r0+(T)(11),
r6 = r0+(T)(13),
r7 = r0+(T)(17);
for(int j=0; j<COMP_ITERATIONS; j+=UNROLL_ITERATIONS){
#pragma unroll
for(int i=0; i<UNROLL_ITERATIONS-memory_ratio; i++){
r0 = mad_op(r0, r0, r4);
r1 = mad_op(r1, r1, r5);
r2 = mad_op(r2, r2, r6);
r3 = mad_op(r3, r3, r7);
r4 = mad_op(r4, r4, r0);
r5 = mad_op(r5, r5, r1);
r6 = mad_op(r6, r6, r2);
r7 = mad_op(r7, r7, r3);
}
bool do_write = true;
int reg_idx = 0;
#pragma unroll
for(int i=UNROLL_ITERATIONS-memory_ratio; i<UNROLL_ITERATIONS; i++){
// Each iteration maps to one memory operation
T& r = reg_idx==0 ? r0 : (reg_idx==1 ? r1 : (reg_idx==2 ? r2 : (reg_idx==3 ? r3 : (reg_idx==4 ? r4 : (reg_idx==5 ? r5 : (reg_idx==6 ? r6 : r7))))));
if( do_write )
data[ array_index+halfarraysize ] = r;
else {
r = data[ array_index ];
if( ++reg_idx>=REGBLOCK_SIZE )
reg_idx = 0;
array_index += index_stride;
}
do_write = !do_write;
}
if( array_index >= array_index_bound ){
if( ++initial_index_factor > initial_index_range)
initial_index_factor = 0;
array_index = index_base + initial_index_factor*index_stride;
}
}
if( (r0==GPU_INF(T)) && (r1==GPU_INF(T)) && (r2==GPU_INF(T)) && (r3==GPU_INF(T)) &&
(r4==GPU_INF(T)) && (r5==GPU_INF(T)) && (r6==GPU_INF(T)) && (r7==GPU_INF(T)) ){ // extremely unlikely to happen
g_data[0] = r0+r1+r2+r3+r4+r5+r6+r7;
}
}
void initializeEvents(hipEvent_t *start, hipEvent_t *stop){
CUDA_SAFE_CALL( hipEventCreate(start) );
CUDA_SAFE_CALL( hipEventCreate(stop) );
CUDA_SAFE_CALL( hipEventRecord(*start, 0) );
}
float finalizeEvents(hipEvent_t start, hipEvent_t stop){
CUDA_SAFE_CALL( hipGetLastError() );
CUDA_SAFE_CALL( hipEventRecord(stop, 0) );
CUDA_SAFE_CALL( hipEventSynchronize(stop) );
float kernel_time;
CUDA_SAFE_CALL( hipEventElapsedTime(&kernel_time, start, stop) );
CUDA_SAFE_CALL( hipEventDestroy(start) );
CUDA_SAFE_CALL( hipEventDestroy(stop) );
return kernel_time;
}
void runbench_warmup(double *cd, long size){
const long reduced_grid_size = size/(UNROLLED_MEMORY_ACCESSES)/32;
const int BLOCK_SIZE = 256;
const int TOTAL_REDUCED_BLOCKS = reduced_grid_size/BLOCK_SIZE;
dim3 dimBlock(BLOCK_SIZE, 1, 1);
dim3 dimReducedGrid(TOTAL_REDUCED_BLOCKS, 1, 1);
hipLaunchKernel(HIP_KERNEL_NAME(benchmark_func< short, BLOCK_SIZE, 0 >), dim3(dimReducedGrid), dim3(dimBlock ), 0, 0, (short)1, (short*)cd);
CUDA_SAFE_CALL( hipGetLastError() );
CUDA_SAFE_CALL( hipDeviceSynchronize() );
}
template<int memory_ratio>
void runbench(double *cd, long size){
if( memory_ratio>UNROLL_ITERATIONS ){
fprintf(stderr, "ERROR: memory_ratio exceeds UNROLL_ITERATIONS\n");
exit(1);
}
const long compute_grid_size = size/(UNROLLED_MEMORY_ACCESSES)/2;
const int BLOCK_SIZE = 256;
const int TOTAL_BLOCKS = compute_grid_size/BLOCK_SIZE;
const long long computations = 2*(long long)(COMP_ITERATIONS)*REGBLOCK_SIZE*compute_grid_size;
const long long memoryoperations = (long long)(COMP_ITERATIONS)*compute_grid_size;
dim3 dimBlock(BLOCK_SIZE, 1, 1);
dim3 dimGrid(TOTAL_BLOCKS, 1, 1);
hipEvent_t start, stop;
initializeEvents(&start, &stop);
hipLaunchKernel(HIP_KERNEL_NAME(benchmark_func< float, BLOCK_SIZE, memory_ratio >), dim3(dimGrid), dim3(dimBlock ), 0, 0, 1.0f, (float*)cd);
float kernel_time_mad_sp = finalizeEvents(start, stop);
initializeEvents(&start, &stop);
hipLaunchKernel(HIP_KERNEL_NAME(benchmark_func< double, BLOCK_SIZE, memory_ratio >), dim3(dimGrid), dim3(dimBlock ), 0, 0, 1.0, cd);
float kernel_time_mad_dp = finalizeEvents(start, stop);
initializeEvents(&start, &stop);
hipLaunchKernel(HIP_KERNEL_NAME(benchmark_func< int, BLOCK_SIZE, memory_ratio >), dim3(dimGrid), dim3(dimBlock ), 0, 0, 1, (int*)cd);
float kernel_time_mad_int = finalizeEvents(start, stop);
const double memaccesses_ratio = (double)(memory_ratio)/UNROLL_ITERATIONS;
const double computations_ratio = 1.0-memaccesses_ratio;
printf(" %4d, %8.3f,%8.2f,%8.2f,%7.2f, %8.3f,%8.2f,%8.2f,%7.2f, %8.3f,%8.2f,%8.2f,%7.2f\n",
UNROLL_ITERATIONS-memory_ratio,
(computations_ratio*(double)computations)/(memaccesses_ratio*(double)memoryoperations*sizeof(float)),
kernel_time_mad_sp,
(computations_ratio*(double)computations)/kernel_time_mad_sp*1000./(double)(1000*1000*1000),
(memaccesses_ratio*(double)memoryoperations*sizeof(float))/kernel_time_mad_sp*1000./(1000.*1000.*1000.),
(computations_ratio*(double)computations)/(memaccesses_ratio*(double)memoryoperations*sizeof(double)),
kernel_time_mad_dp,
(computations_ratio*(double)computations)/kernel_time_mad_dp*1000./(double)(1000*1000*1000),
(memaccesses_ratio*(double)memoryoperations*sizeof(double))/kernel_time_mad_dp*1000./(1000.*1000.*1000.),
(computations_ratio*(double)computations)/(memaccesses_ratio*(double)memoryoperations*sizeof(int)),
kernel_time_mad_int,
(computations_ratio*(double)computations)/kernel_time_mad_int*1000./(double)(1000*1000*1000),
(memaccesses_ratio*(double)memoryoperations*sizeof(int))/kernel_time_mad_int*1000./(1000.*1000.*1000.) );
}
extern "C" void mixbenchGPU(double *c, long size){
const char *benchtype = "compute with global memory (block strided)";
printf("Trade-off type: %s\n", benchtype);
double *cd;
CUDA_SAFE_CALL( hipMalloc((void**)&cd, size*sizeof(double)) );
// Copy data to device memory
CUDA_SAFE_CALL( hipMemset(cd, 0, size*sizeof(double)) ); // initialize to zeros
// Synchronize in order to wait for memory operations to finish
CUDA_SAFE_CALL( hipDeviceSynchronize() );
printf("---------------------------------------------------------- CSV data ----------------------------------------------------------\n");
printf("Experiment ID, Single Precision ops,,,, Double precision ops,,,, Integer operations,,, \n");
printf("Compute iters, Flops/byte, ex.time, GFLOPS, GB/sec, Flops/byte, ex.time, GFLOPS, GB/sec, Iops/byte, ex.time, GIOPS, GB/sec\n");
runbench_warmup(cd, size);
runbench<32>(cd, size);
runbench<31>(cd, size);
runbench<30>(cd, size);
runbench<29>(cd, size);
runbench<28>(cd, size);
runbench<27>(cd, size);
runbench<26>(cd, size);
runbench<25>(cd, size);
runbench<24>(cd, size);
runbench<23>(cd, size);
runbench<22>(cd, size);
runbench<21>(cd, size);
runbench<20>(cd, size);
runbench<19>(cd, size);
runbench<18>(cd, size);
runbench<17>(cd, size);
runbench<16>(cd, size);
runbench<15>(cd, size);
runbench<14>(cd, size);
runbench<13>(cd, size);
runbench<12>(cd, size);
runbench<11>(cd, size);
runbench<10>(cd, size);
runbench<9>(cd, size);
runbench<8>(cd, size);
runbench<7>(cd, size);
runbench<6>(cd, size);
runbench<5>(cd, size);
runbench<4>(cd, size);
runbench<3>(cd, size);
runbench<2>(cd, size);
runbench<1>(cd, size);
runbench<0>(cd, size);
printf("---------------------------------------------------------- CSV data ----------------------------------------------------------\n");
// Copy results back to host memory
CUDA_SAFE_CALL( hipMemcpy(c, cd, size*sizeof(double), hipMemcpyDeviceToHost) );
CUDA_SAFE_CALL( hipFree(cd) );
CUDA_SAFE_CALL( hipDeviceReset() );
}