/micro_benchmarks: Contains all the microbenchmark tests we run on GPU hardware to get
latency, effective bandwiths, and effective flops compuations. Contains README.txt
describing more in detail the benchmarks and how to compile and run.
/benchmarks: Contains all the benchmarks mentioned in our paper such as
N-body, 2D-Stencil, Conjugate Gradient, and D2D. Each benchmark contains instructions
in how to compile, run, and contains python script to generate plots
/Data: Contains all csv file from our complete benchmark runs
/Plots: Contains all our plots
predictiveModel.py: Contains python code of our predictive model, Reading csv data, and generates plots plothepers.py: Helps makes pretty plots
All micro_benchmarks and benchmarks we completed on the CIMS machine cuda3 on cuda libary 12.6