Parallel Programming with NVIDIA CUDA

Using hardware acceleration via General Programming on stock GPUs (GPGPU), I've sped up my algorithms by more than tenfold. This article shows how you can achieve these results too!

If parallelization of your algorithm is possible, using CUDA will speed up your computations dramatically, allowing you to make the most out of your hardware.

The main challenge consists in deciding how to partition your problem into chunks suitable for parallel execution. As with so many other aspects in parallel programming, this is where experience and—why not—imagination come into play.

Additional techniques offer room for even more improvement. In particular, the on-chip shared memory of each compute node allows further speedup of the computation process.

Alejandro Segovia is a parallel programming advisor for CoroWare. He is also a contributing partner at RealityFrontier. He works in 3-D graphic development and GPU acceleration. Alejandro was recently a visiting scientist at the University of Delaware where he investigated CUDA from an academic standpoint. His findings were published at the IEEE IPCCC Conference in 2009.



Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

The statement minima[y][x] =

Anonymous's picture

The statement
minima[y][x] = (norm(field[y][x]) < threshold) ? true : false
may incur branching penalty

You can just use the first part
minima[y][x] = (norm(field[y][x]) < threshold)