SIMD_AVX_Example/README.md

Performance Test

Example Project to test the performance increase of avx multiplication in contrast to standard multiplication by multiplying 16 32bit integer a fixed number of times.

Programstructure

The program creates two arrays with each containing 16 integers which are dispatched to a thread in a timer environment to track the execution time. The results are then printed to stdout.

Benchmark by 2147483647 iterations

	./program	nice -20 program	chrt -f 99 program
AVX	9.46229s	6.651s	11.7455s
Std	16.7001s	13.6682s	16.828s