benchmarks/blas.cpp
/*******************************************************
* Copyright (c) 2014, ArrayFire
* All rights reserved.
*
* This file is distributed under 3-clause BSD license.
* The complete license agreement can be obtained at:
* https://arrayfire.com/licenses/BSD-3-Clause
********************************************************/
#include <arrayfire.h>
#include <math.h>
#include <stdio.h>
#include <cstdlib>
#include <string>
using namespace af;
// create a small wrapper to benchmark
static array A; // populated before each timing
static void fn() {
array B = matmul(A, A); // matrix multiply
}
int main(int argc, char** argv) {
double peak = 0;
try {
int device = argc > 1 ? atoi(argv[1]) : 0;
setDevice(device);
const std::string dtype(argc > 2 ? argv[2] : "f32");
const af_dtype dt = (dtype == "f16" ? f16 : f32);
if (dt == f16)
printf("Device %d isHalfAvailable ? %s\n", device,
isHalfAvailable(device) ? "yes" : "no");
info();
printf("Benchmark N-by-N matrix multiply at %s \n", dtype.c_str());
for (int n = 128; n <= 2048; n += 128) {
printf("%4d x %4d: ", n, n);
A = constant(1, n, n, dt);
double time = timeit(fn); // time in seconds
double gflops = 2.0 * powf(n, 3) / (time * 1e9);
if (gflops > peak) peak = gflops;
printf(" %4.0f Gflops\n", gflops);
fflush(stdout);
}
} catch (af::exception& e) {
fprintf(stderr, "%s\n", e.what());
throw;
}
printf(" ### peak %g GFLOPS\n", peak);
return 0;
}
af::matmul
AFAPI array matmul(const array &lhs, const array &rhs, const matProp optLhs=AF_MAT_NONE, const matProp optRhs=AF_MAT_NONE)
Matrix multiply of two arrays.
af::isHalfAvailable
AFAPI bool isHalfAvailable(const int device)
Queries the current device for half precision floating point support.
af::info
AFAPI void info()
af::constant
array constant(T val, const dim4 &dims, const dtype ty=(af_dtype) dtype_traits< T >::ctype)
af::setDevice
AFAPI void setDevice(const int device)
Sets the current device.
af::dtype
af_dtype dtype
Definition: defines.h:549
af::array
A multi dimensional data container.
Definition: array.h:35
af
Definition: algorithm.h:15
af::timeit
AFAPI double timeit(void(*fn)())
af_dtype
af_dtype
Definition: defines.h:210
f16
@ f16
16-bit floating point value
Definition: defines.h:228
af::exception
An ArrayFire exception class.
Definition: exception.h:29
arrayfire.h
af::exception::what
virtual const char * what() const
Returns an error message for the exception in a string format.
Definition: exception.h:60
f32
@ f32
32-bit floating point values
Definition: defines.h:211