APSP-CUDA is a software implementation of the All-Pairs Shortest Path algorithm on a GPU using CUDA.
It is important to note that Octave code was written as a wrapper, in order to take the time measurements easier.
Read APSP-CUDA_report.pdf for more information.
It was a project for the lesson "Parallel & Distributed Systems" by prof. Nikos P. Pitsianis, at Aristotle University of Thessaloniki in 2016.
You can contact me by email: Marios Mitalidis - mmitalidis@gmail.com