Lesson 3 - Fundamental GPU Algorithms(Reduce,Scan,Histogram)
目录
8.练习:Binary and Associative Operators
9.练习:Serial lmplementation of Reduce
11.练习:Step Complexity of Parallel Reduce
14.Balancing Checkbook With Scan
17.Why Do We Care About Parallel Scan
18.Inclusive Vs. Exclusive Scan
19.练习:Serial Implementation of Scan
20.练习:Inclusive Scan Revisited
21.Hills Steele vs Blelloch Scan
24.Analyzing The Blelloch Downsweep
25.练习:Which Scan Should You Use
27.练习:Serial Implementation of Histogram
28.Parallel Implementation of Histogram
29.练习:Implementing Histogram Using Atomics
30.Implementing Histogram Using Local Memory
31.Calculating Global Histogram Using Reduction
33.练习:What Operation should we use
34.练习:Final Thoughts on Histogram
35.Description of Problem Set #3
1.Welcome to Unit#3
2.Fundamental GPU Algorithms
回顾Lession2的 communication pattern
3.Digging Holes Again
4.Steps and Work
5.练习:A Quiz on Step and work
6.Reduce Part 1
7.Reduce Part 2
8.练习:Binary and Associative Operators
9.练习:Serial lmplementation of Reduce
10.练习:Parallel Reduce
11.练习:Step Complexity of Parallel Reduce
12.练习:Reduction Using Global and Shared Memory
#include <stdio.h>
#include <stdlib.h>
#include <cuda_runtime.h>
__global__ void global_reduce_kernel(float * d_out, float * d_in)
{
int myId = threadIdx.x + blockDim.x * blockIdx.x;
int tid = threadIdx.x;
// do reduction in global mem
for (unsigned int s = blockDim.x / 2; s > 0; s >>= 1)
{
if (tid < s)
{
d_in[myId] += d_in[myId + s];
}
__syncthreads(); // make sure all adds at one stage are done!
}
// only thread 0 writes result for this block back to global mem
if (tid == 0)
{
d_out[blockIdx.x] = d_in[myId];
}
}
__global__ void shmem_reduce_kernel(float * d_out, const float * d_in)
{
// sdata is allocated in the kernel call: 3rd arg to <<<b, t, shmem>>>
extern __shared__ float sdata[];
int myId = threadIdx.x + blockDim.x * blockIdx.x;
int tid = threadIdx.x;
// load shared mem from global mem
sdata[tid] = d_in[myId];
__syncthreads(); // make sure entire block is loaded!
// do reduction in shared mem
for (unsigned int s = blockDim.x / 2; s > 0; s >>= 1)
{
if (tid < s)
{
sdata[tid] += sdata[tid + s];
}
__syncthreads(); // make sure all adds at one stage are done!
}
// only thread 0 writes result for this block back to global mem
if (tid == 0)
{
d_out[blockIdx.x] = sdata[0];
}
}
void reduce(float * d_out, float * d_intermediate, float * d_in,
int size, bool usesSharedMemory)
{
// assumes that size is not greater than maxThreadsPerBlock^2
// and that size is a multiple of maxThreadsPerBlock
const int maxThreadsPerBlock = 1024;
int threads = maxThreadsPerBlock;
int blocks = size / maxThreadsPerBlock;
if (usesSharedMemory)
{
shmem_reduce_kernel<<<blocks, threads, threads * sizeof(float)>>>
(d_intermediate, d_in);
}
else
{
global_reduce_kernel<<<blocks, threads>>>
(d_intermediate, d_in);
}
// now we're down to one block left, so reduce it
threads = blocks; // launch one thread for each block in prev step
blocks = 1;
if (usesSharedMemory)
{
shmem_reduce_kernel<<<blocks, threads, threads * sizeof(float)>>>
(d_out, d_intermediate);
}
else
{
global_reduce_kernel<<<blocks, threads>>>
(d_out, d_intermediate);
}
}
int main(int argc, char **argv)
{
int deviceCount;
cudaGetDeviceCount(&deviceCount);
if (deviceCount == 0) {
fprintf(stderr, "error: no devices supporting CUDA.\n");
exit(EXIT_FAILURE);
}
int dev = 0;
cudaSetDevice(dev);
cudaDeviceProp devProps;
if (cudaGetDeviceProperties(&devProps, dev) == 0)
{
printf("Using device %d:\n", dev);
printf("%s; global mem: %dB; compute v%d.%d; clock: %d kHz\n",
devProps.name, (int)devProps.totalGlobalMem,
(int)devProps.major, (int)devProps.minor,
(int)devProps.clockRate);
}
const int ARRAY_SIZE = 1 << 20;
const int ARRAY_BYTES = ARRAY_SIZE * sizeof(float);
// generate the input array on the host
float h_in[ARRAY_SIZE];
float sum = 0.0f;
for(int i = 0; i < ARRAY_SIZE; i++) {
// generate random float in [-1.0f, 1.0f]
h_in[i] = -1.0f + (float)random()/((float)RAND_MAX/2.0f);
sum += h_in[i];
}
// declare GPU memory pointers
float * d_in, * d_intermediate, * d_out;
// allocate GPU memory
cudaMalloc((void **) &d_in, ARRAY_BYTES);
cudaMalloc((void **) &d_intermediate, ARRAY_BYTES); // overallocated
cudaMalloc((void **) &d_out, sizeof(float));
// transfer the input array to the GPU
cudaMemcpy(d_in, h_in, ARRAY_BYTES, cudaMemcpyHostToDevice);
int whichKernel = 0;
if (argc == 2) {
whichKernel = atoi(argv[1]);
}
cudaEvent_t start, stop;
cudaEventCreate(&start);
cudaEventCreate(&stop);
// launch the kernel
switch(whichKernel) {
case 0:
printf("Running global reduce\n");
cudaEventRecord(start, 0);
for (int i = 0; i < 100; i++)
{
reduce(d_out, d_intermediate, d_in, ARRAY_SIZE, false);
}
cudaEventRecord(stop, 0);
break;
case 1:
printf("Running reduce with shared mem\n");
cudaEventRecord(start, 0);
for (int i = 0; i < 100; i++)
{
reduce(d_out, d_intermediate, d_in, ARRAY_SIZE, true);
}
cudaEventRecord(stop, 0);
break;
default:
fprintf(stderr, "error: ran no kernel\n");
exit(EXIT_FAILURE);
}
cudaEventSynchronize(stop);
float elapsedTime;
cudaEventElapsedTime(&elapsedTime, start, stop);
elapsedTime /= 100.0f; // 100 trials
// copy back the sum from GPU
float h_out;
cudaMemcpy(&h_out, d_out, sizeof(float), cudaMemcpyDeviceToHost);
printf("average time elapsed: %f\n", elapsedTime);
// free GPU memory allocation
cudaFree(d_in);
cudaFree(d_intermediate);
cudaFree(d_out);
return 0;
}
[email protected]:~/Desktop/cs344-master/Lesson Code Snippets/Lesson 3 Code Snippets$ ./a.out 0
Using device 0:
GeForce GTX 980 Ti; global mem: 2083127296B; compute v5.2; clock: 1228000 kHz
Running global reduce
average time elapsed: 0.138348
[email protected]:~/Desktop/cs344-master/Lesson Code Snippets/Lesson 3 Code Snippets$ ./a.out 1
Using device 0:
GeForce GTX 980 Ti; global mem: 2083127296B; compute v5.2; clock: 1228000 kHz
Running reduce with shared mem
average time elapsed: 0.117061
13.Scan
14.Balancing Checkbook With Scan
15.练习:Inputs to Scan
16.练习:What Scan Actually Does
17.Why Do We Care About Parallel Scan
18.Inclusive Vs. Exclusive Scan
19.练习:Serial Implementation of Scan
#include <stdio.h>
int main(int argc,char **argv)
{
const int ARRAY_SIZE = 10;
int acc = 0;
int out[ARRAY_SIZE];
int elements[] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10};
for(int i = 0; i < ARRAY_SIZE; i++){
acc = acc + elements[i];
out[i] = acc;
}
for(int i = 0 ; i < ARRAY_SIZE; i++){
printf("%i ", out[i]);
}
return 0;
}
20.练习:Inclusive Scan Revisited
第n次输出需要加n-1次 0 +1 +2 +...+ n-1
串行的step需要走O(n)步
21.Hills Steele vs Blelloch Scan
22.练习:Hills Steele Scan