SIMD matrix multiplication analysis

Computer Structure and Language optional assembly project

Project No.3: Using SIMD in x86_64 assembly to optimize matrix multiplication

Usage

Each shell script takes a specific type of file (either C code or asm) which includes matrix_mul function as an input and compiles it and links it with timer.c
timer.c will run the linked function multiple times according to defined constants and outputs the results as binary in results_*.bin
plot.py will read the results from the files specified in the code and plots them accordingly.
To run and give input to files, go to runnable_code and simply either compile the C code or use run-asm.sh [filename] to compile and run the assembly codes.

Name		Name	Last commit message	Last commit date
Latest commit History 44 Commits
data_1024		data_1024
data_2048		data_2048
runnable_code		runnable_code
.gitignore		.gitignore
README.md		README.md
analysis.pdf		analysis.pdf
asm_matrix_mul.s		asm_matrix_mul.s
asm_matrix_mul_cache_efficient.s		asm_matrix_mul_cache_efficient.s
asm_time.sh		asm_time.sh
c_matrix_mul.c		c_matrix_mul.c
c_matrix_mul_cache_efficient.c		c_matrix_mul_cache_efficient.c
description.pdf		description.pdf
linux_time.sh		linux_time.sh
linux_time_O3.sh		linux_time_O3.sh
plot.py		plot.py
simd_matrix_mul.s		simd_matrix_mul.s
timer.c		timer.c
win_time.bat		win_time.bat