K20s lead to scientific breakthrough

Good to hear!

http://www.eweek.com/innovation/nvidia-gpu-cray-supercomputer-power-hiv-research-breakthrough/

At my work we are using them(K20s) to interpret and analyse large EEG data sets, so hopefully someday we will contribute towards some advancements in the treatment of Dementia and Alzheimers disease.

Is there a published paper about the project you are working on?

My role so far has been to convert existing MATLAB code to faster CUDA versions. The researchers I work with have been ‘first-names’ on papers related to EEG analysis, but nothing yet directly related to this batch of current 128 channel EEG work.

Soon a new paper on a conic optimization problem solver will be published, and there will be a link to my CUDA version of the solver(in addition to the author’s MATLAB and C version). I will link to that once it is published.

There was a Nvidia spotlight on my lab (The majority of the CUDA convex optimization work was mine);

http://www.nvidia.com/content/cuda/spotlights/adam-gazzaley-ucsf.html

Still more to do, but thanks to Nvidia we have made much progress on our neuroscience work.