Authors: Alexiane Laude, Arthur Pignet
The aim of this project is to implement a vectorized and multi-threaded function in C, to compute as fast as possible (in double precision) the following sum:
for a vector
To compile the project using gcc, clone this repository
git clone https://github.com/alexianelaude/projet-archi.git
cd projet-archi
and then run:
gcc -lpthread -mavx2 -O3 -o project project.c
You can finally execute project.exe
VALEURS
Sequentiel (scalaire: 6.991246e+005, vectoriel: 6.991246e+005) Parallele (nb_threads: 8, scalaire: 6.991246e+005, vectoriel: 6.991246e+005)
TEMPS D EXECUTION
Sequentiel (scalaire: 6.598651e-003, vectoriel: 1.093769e-003) Parallele (nb_threads: 8, scalaire: 2.031269e-003, , vectoriel: 7.812715e-004)
Acceleration (vectoriel: 6.032948e+000, multithread: 3.248536e+000, vectoriel + multithread : 8.446042e+000)