Speeding Up Compressed Sensing Algorithms

Are you looking for ways to speed up compressed sensing? If you work in the areas of medical image reconstruction, image acquisition or sensor networks, you probably are. This paper, Parallel Implementation of Compressed Sensing Algorithm on CUDA-GPU, compares CPUs running Matlab and GPUs running Jacket using a Basis Pursuit Algorithm for compressed sensing.

Digital Holograms Faster than Ever

REAL3D is a digital holography project funded by the EU and brings together nine participants from academia and industry under the FP7.

Feature Learning Architectures with GPU-acceleration

Stanford researchers in Andrew Ng’s group used GPUs and Jacket to speed up their work on Feature Learning Architectures. They wanted to know why certain feature learning architectures with random, untrained weights perform so well on object recognition tasks. The complete write up can be found here.

Improved Fat/Water Reconstruction Algorithm with Jacket

Learn how Case Western Reserve University researchers turned to GPUs running Jacket to develop a fast and robust Iterative Decomposition of water and fat with an Echo Asymmetry and Least-squares (IDEAL) reconstruction algorithm.  Read more:

Back and Forth Error Compensation and Correction (BFECC)

I recently came across Song at al.'s Stable But Nondissipative Water and eventually with the Back and Forth Error Compensation and Correction technique. I must say the simulation with BFECC looks great (looks less dissipated, with more small features preservation) and it has almost no impact in performance and can be trivially added to one's current implementation. It also has the advantage of being very intuitive.


A couple of implementation notes:

Registration call for OpenCL workshop Brisbane on 1st December 2010

Registration call for OpenCL workshop Brisbane on 1st December 2010

F# Black-Scholes running on GPIs and SSE3 Multicore Processors using Accelerator

I've written an article illustrating how to implement the Black-Scholes option pricing algorithm using Microsoft's Accelerator GPGPU system (which runs on any card that supports DX9). The link to the article is shown below and it contains a link to the download page for Accelerator:

Problems with texture memory and shared memory (don't know where)

I use the texture memory like this:

texture<int, 2, cudaReadModeElementType> texType;

texType.addressMode[0] = cudaAddressModeClamp;

texType.addressMode[1] = cudaAddressModeClamp;

texType.filterMode = cudaFilterModePoint;

texType.normalized = false;

cudaBindTextureToArray (texType, typeCuArray, channelDesc_V);

SiteType temp2 = (SiteType)tex2D(texType, index%widthTex, index/widthTex);

and then:

X[tdx] = temp2;

where X is shared memory:

Torben’s Corner – A GPU Computing Gem for Jacket Programmers!

In January, we introduced you to Torben’s Corner – a resource wiki created and maintained by Jacket programming guru, Torben Larsen at Aalborg University in Denmark.  Many Jacket programmers have gained valuable insights from Torben’s Corner, including GPU performance charts, coding guidelines, special tricks.


Subscribe to RSS - blogs