matrusse

Fast matrix multiplication other GF2 using 4-russians algorithm

Requirement

• cmake version 3.14
• a c compiler, tested with gcc v8.3.1
• a c++ compiler

Compilation

• clone the repository and go inside : git clone https://www.github.com/Drapsag74/matrusse.git && cd matrusse
• create a new directory for your binary for example : mkdir bin && cd bin
• now you can generate the make file by running cmake -DCMAKE_BUILD_TYPE=Release ..
• then you can build the binaries by running make all, or make Matrusse if you only want the Matrusse binary
• binaries are located in src

Execution

• To run the main program, you have to specify which algorithm you want to try. Then for every algorithm, you have to set the proper number of parameters.
• -b to run the filltable algorithms (1 for fillTable, 2 for fillTable2)
• -m to run the matrusse algorithms (1 for matrusseV1, 2 for matrusseV2TestBloc, 3 for matrusseIntrin, 4 for matrusseV2SpeedOfLight and 5 for matrusseV2_1TestBloc)
• -t to run the thread algorithm (not finished)
• for example, if you want to do the multiplication of a matrix A (1000x500) with a matrix B (500x1000) with the algorithm "matrusseintrin" and a block of size 8 : ./Matrusse 1000 500 8 -m3