chizeta/cnorxz
chizeta/cnorxz
0
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

make comparison loop absolutely equivalent to ma test loop -> there is no notable difference in the performance anymore

Browse source
This commit is contained in:
Christian Zimmermann 2019-02-27 19:02:06 +01:00
parent 09f8074910
commit 25fadae974
1 changed files with 27 additions and 23 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

50
src/tests/op_perf_test.cc
View file

@ -9,6 +9,7 @@
#include <ctime>
#include <cmath>
#include <chrono>
#define ONLY_SPIN
@ -135,7 +136,7 @@ namespace {
typedef MultiRangeFactory<SR,SR,SR,SR,SR,SR,SR,SR> SR8F;
typedef SR8F::oType SR8;
static const size_t os = 300;
static const size_t os = 3000;
static const size_t is = 65536;
static const size_t s = is*os;
@ -176,17 +177,9 @@ namespace {
//auto deltap = MAT::getIndex<SR>();
auto deltap = MAT::getIndex<GenSingleRange<size_t,SpaceType::NONE,1>>();
auto mix = MAT::mkMIndex( alpha, beta, gamma, jj );
auto mix = MAT::mkMIndex( jj, alpha, beta, gamma );
std::clock_t begin = std::clock();
//for(size_t i = 0; i != os; ++i){
//res1(ii ,delta, deltap).par() += ma(ii, delta, alpha, alpha, beta, beta, gamma, gamma, deltap).c(mix);
tcast<v256>(res1)(ii ,delta, deltap) += tcast<v256>(ma)(delta, alpha, alpha, beta, beta, gamma, gamma, deltap).c(mix);
//}
std::clock_t end = std::clock();
std::cout << "MultiArray time: " << static_cast<double>( end - begin ) / CLOCKS_PER_SEC
<< std::endl;
vector<double> vres(4*4*os);
for(size_t d = 0; d != 4; ++d){
for(size_t p = 0; p != 4; ++p){
@ -194,16 +187,15 @@ namespace {
vres[tidx] = 0.;
}
}
std::clock_t begin2 = std::clock();
auto begin2 = std::chrono::system_clock::now();
double* vrptr = vres.data();
double* dptr = data.data();
for(size_t j = 0; j != os; ++j) {
for(size_t i = 0; i != os; ++i){
for(size_t a = 0; a != 4; ++a){
for(size_t b = 0; b != 4; ++b){
for(size_t c = 0; c != 4; ++c){
for(size_t d = 0; d != 4; ++d){
for(size_t i = 0; i != os; ++i){
for(size_t d = 0; d != 4; ++d){
for(size_t j = 0; j != os; ++j) {
for(size_t a = 0; a != 4; ++a){
for(size_t b = 0; b != 4; ++b){
for(size_t c = 0; c != 4; ++c){
const size_t tidx = i*4*4 + d*4;
const size_t sidx = /*i*65536 +*/ d*4*4*4*4*4*4*4 + a*5*4*4*4*4*4 + b*5*4*4*4 + c*5*4;
double* xvrptr = vrptr + tidx;
@ -218,8 +210,20 @@ namespace {
}
}
}
std::clock_t end2 = std::clock();
auto end2 = std::chrono::system_clock::now();
std::cout << "vector - for loop time: " << std::chrono::duration<double>(end2-begin2).count()
<< std::endl;
auto begin = std::chrono::system_clock::now();
//for(size_t i = 0; i != os; ++i){
//res1(ii ,delta, deltap).par() += ma(ii, delta, alpha, alpha, beta, beta, gamma, gamma, deltap).c(mix);
tcast<v256>(res1)(ii ,delta, deltap).par() += tcast<v256>(ma)(delta, alpha, alpha, beta, beta, gamma, gamma, deltap).c(mix);
//}
auto end = std::chrono::system_clock::now();
std::cout << "MultiArray time: " << std::chrono::duration<double>(end-begin).count()
<< std::endl;
assert( xround(res1.at(mkts(0,0,0))) == xround(vres[0]) );
assert( xround(res1.at(mkts(0,0,1))) == xround(vres[1]) );
assert( xround(res1.at(mkts(0,0,2))) == xround(vres[2]) );
@ -240,9 +244,9 @@ namespace {
assert( xround(res1.at(mkts(0,3,2))) == xround(vres[14]) );
assert( xround(res1.at(mkts(0,3,3))) == xround(vres[15]) );
std::cout << "vector - for loop time: " << static_cast<double>( end2 - begin2 ) / CLOCKS_PER_SEC
<< std::endl;
std::cout << "ratio: " << static_cast<double>( end - begin ) / static_cast<double>( end2 - begin2 ) << std::endl;
std::cout << "ratio: "
<< std::chrono::duration<double>(end-begin).count() / std::chrono::duration<double>(end2-begin2).count()
<< std::endl;
}
#ifndef ONLY_SPIN
void OpTest_Performance::PCheck()