GPU Spam Filter
The project was part of “Project in Parallel/Distributed Programming” class in the Technion. In this project we implemented a Bayesian spam filter on a GPU using CUDA.
A Framework for Auto-Parallelization of Stencil Computations on a Cluster of GPUs or CPUs
The project was part of “Project in Parallel/Distributed Programming” class in the Technion. In this project I implemented a framework that can perform any stencil computation, where the user chooses the stencil function. The framework distributes the computation using MPI while each CPU can use a GPU to speed up its computation.
