detectors/aare
25
0
Fork 0
mirror of https://github.com/slsdetectorgroup/aare.git synced 2025-06-15 16:57:12 +02:00
Code Issues Actions Packages Releases Activity

formatted main branch

Browse Source
This commit is contained in:
mazzol_a
2025-06-10 16:09:06 +02:00
parent efd2338f54
commit f9751902a2
87 changed files with 1710 additions and 1639 deletions
Download Patch File Download Diff File Expand all files Collapse all files

210
benchmarks/ndarray_benchmark.cpp
View File

@ -1,136 +1,132 @@
#include <benchmark/benchmark.h>
#include "aare/NDArray.hpp"
#include <benchmark/benchmark.h>
using aare::NDArray;
constexpr ssize_t size = 1024;
class TwoArrays : public benchmark::Fixture {
public:
NDArray<int,2> a{{size,size},0};
NDArray<int,2> b{{size,size},0};
void SetUp(::benchmark::State& state) {
for(uint32_t i = 0; i < size; i++){
for(uint32_t j = 0; j < size; j++){
a(i, j)= i*j+1;
b(i, j)= i*j+1;
}
public:
NDArray<int, 2> a{{size, size}, 0};
NDArray<int, 2> b{{size, size}, 0};
void SetUp(::benchmark::State &state) {
for (uint32_t i = 0; i < size; i++) {
for (uint32_t j = 0; j < size; j++) {
a(i, j) = i * j + 1;
b(i, j) = i * j + 1;
}
}
}
}
// void TearDown(::benchmark::State& state) {
// }
// void TearDown(::benchmark::State& state) {
// }
};
BENCHMARK_F(TwoArrays, AddWithOperator)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res = a+b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, AddWithIndex)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) + b(i);
BENCHMARK_F(TwoArrays, AddWithOperator)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res = a + b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, AddWithIndex)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) + b(i);
}
benchmark::DoNotOptimize(res);
}
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, SubtractWithOperator)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res = a-b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, SubtractWithIndex)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) - b(i);
BENCHMARK_F(TwoArrays, SubtractWithOperator)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res = a - b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, SubtractWithIndex)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) - b(i);
}
benchmark::DoNotOptimize(res);
}
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, MultiplyWithOperator)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res = a*b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, MultiplyWithIndex)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) * b(i);
BENCHMARK_F(TwoArrays, MultiplyWithOperator)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res = a * b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, MultiplyWithIndex)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) * b(i);
}
benchmark::DoNotOptimize(res);
}
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, DivideWithOperator)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res = a/b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, DivideWithIndex)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) / b(i);
BENCHMARK_F(TwoArrays, DivideWithOperator)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res = a / b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, DivideWithIndex)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) / b(i);
}
benchmark::DoNotOptimize(res);
}
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, FourAddWithOperator)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res = a+b+a+b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, FourAddWithIndex)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) + b(i) + a(i) + b(i);
BENCHMARK_F(TwoArrays, FourAddWithOperator)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res = a + b + a + b;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, FourAddWithIndex)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) + b(i) + a(i) + b(i);
}
benchmark::DoNotOptimize(res);
}
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, MultiplyAddDivideWithOperator)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res = a*a+b/a;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, MultiplyAddDivideWithIndex)(benchmark::State& st) {
for (auto _ : st) {
// This code gets timed
NDArray<int,2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) * a(i) + b(i) / a(i);
BENCHMARK_F(TwoArrays, MultiplyAddDivideWithOperator)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res = a * a + b / a;
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_F(TwoArrays, MultiplyAddDivideWithIndex)(benchmark::State &st) {
for (auto _ : st) {
// This code gets timed
NDArray<int, 2> res(a.shape());
for (uint32_t i = 0; i < a.size(); i++) {
res(i) = a(i) * a(i) + b(i) / a(i);
}
benchmark::DoNotOptimize(res);
}
benchmark::DoNotOptimize(res);
}
}
BENCHMARK_MAIN();