CUDA Week in Review Newsletter Homepage
Mon., Mar. 26, 2012, Issue #72 - Newsletter Home
Welcome to CUDA: WEEK IN REVIEW, an online news summary for the worldwide CUDA, GPGPU and parallel programming ecosystem.
GPU Technology Conference


NVIDIA is looking for smart, passionate software engineers for its high-performance computing team. Relevant domains include geosciences, life sciences, computational fluid dynamics, computational chemistry, computational physics, computational finance, data mining, medical imaging and more. Learn more:


New Parallel Programming Blog

Mark Harris, Chief Technologist for GPU Computing at NVIDIA, has launched a new parallel programming blog called Parallel Forall. Mark was one of the pioneers of GPU computing; he helped kick-start the field while getting his PhD at UNC in 2003 and has been driving the application of GPUs ever since. He will write about technical parallel programming topics and will occasionally invite guest experts from industry and academia to present their techniques and perspectives.
- Read Parallel Forall
New Parallel Programming Blog


Inside Stanford's High-Performance GPU Cluster

This week's CUDA Spotlight is on Brian Tempero of Stanford University's Institute for Computational and Mathematical Engineering (ICME). Brian manages the ICME High Performance GPU Cluster for students and researchers across a range of domains and departments. Brain Tempero
- Read our interview with Brian Tempero

(To suggest a CUDA Spotlight, email:


Rob Farber on "Hot-Rodding Windows and Linux App Performance"

In this Dr. Dobb's article, GPU computing expert Rob Farber explains how "adding GPU capabilities to existing Windows and Linux apps can be done simply using plugins and the built-in support found in CUDA. This easy form of dynamic loading enables CUDA to be used selectively to hugely accelerate individual tasks within a larger application."
- Read the Dr. Dobb's article
- Learn about Rob Farber's CUDA book

Large Scale Machine Learning and CUDA

Adnan Boz provides a thoughtful perspective on running machine learning algorithms with huge amounts of data on the GPU. Note: Adnan is the leader of the South Florida GPU Meetup.
- Read Adnan Boz's blog post

First PGI Compilers with Support for OpenACC Now Available

The Portland Group announced availability of the initial release of its PGI Accelerator Fortran and C compilers with partial support for the new OpenACC spec for directive-based programming of GPUs and accelerators. In the next few months, PGI plans to include additional OpenACC features with full support for the OpenACC 1.0 specs scheduled for release in June.
- Learn more about PGI Compilers

New MATLAB Release Adds GPU Computing Features

The new version of Parallel Computing Toolbox adds features for GPU computing, including support for the full family of FFT functions for all syntaxes; more GPU-enabled MATLAB functions and improved arrayfun function; and asynchronous calculations on the GPU, so that MATLAB continues execution while the GPU runs its calculations at the same time.
- More info on GPU computing in MATLAB
- Special pricing on GPU workstations from HP
- Desktop Engineering article about MATLAB

New Version of SpeedIT from Vratis

SpeedIT is a CUDA-based library for sparse linear algebra that can be also used to accelerate CFD with OpenFOAM.
- Learn about SpeedIT 2.0
- See performance results and comparisons

GTC 2012 Registration is Open

The GPU Technology Conference is on May 14-17. CUDA newsletter readers can enter GM10CD for a 10% discount. (Government/academia eligible for 40% discount with GA40CF.)
- Register for GTC 2012 today


Graduate Fellowship Program for GPU Research, by Chandra Cheij
Supercomputers Power Social Networks, Cancer Research, by Ken Brown
Tesla GPUs Crank Intel Sandy Bridge CPUs To '11', by Sumit Gupta
Our Newest CUDA Centers of Excellence, by Chandra Cheij


Simbios, the Center for Physics-based Simulation of Biological Structures at Stanford University is looking for a software engineer with expertise in molecular dynamics simulations and high performance scientific computing to support and extend the functionality of the OpenMM and Folding@home software projects. Learn more at:


back to the top
Welcome to the new GPU Meetups that recently launched in Pune, India and Northern Colorado!
New York | Silicon Valley | Boston | Denver/Boulder | South Florida  | New Mexico | Austin | Washington, DC | Northern Colorado

Paris | London | South Africa

Tokyo | Perth | Brisbane | Melbourne | New Zealand | India

(Interested in starting a Meetup? Email:


back to the top
- Accelerating Computational Science Symposium (ACSS)
March 28-30, 2012, Washington, DC
Hosted by Oak Ridge Leadership Computing Facility (OLCF), with National Center for Supercomputing Applications (NCSA) and Swiss National Supercomputing Centre (CSCS)

- Accelerating Apps with OpenACC
April 18-19, 2012, Pittsburgh, Pennsylvania
Hosted by Pittsburgh Supercomputing Center in conjunction with PGI and NVIDIA

- SPIE: Defense, Security and Sensing
April 23-27, 2012, Baltimore, Maryland
Note: GPU for Defense Apps talk, April 25, Eric Kelmelis, EM Photonics

- State-of-the-Art Algorithms for Molecular Dynamics
April 30-May 2, 2012, Edinburgh, UK

- INPAR 2012
May 13-14, 2012, San Jose, Calif.

- GPU Technology Conference (GTC 2012)
May 14-17, 2012, San Jose, Calif.

- Intro to GPGPU and CUDA (CINECA)
May 24-25, 2012, Bologna, Italy

- Workshop on Emerging Parallel Architectures (WEPA 2012)
June 2-4, 2012, Omaha, Nebraska

- ISC '12
June 17-21, 2012, Hamburg, Germany

- Parallel Computing Summer School (CINECA)
July 2-13, 2012 (in Italian)
Sept. 3-14, 2012 (in English)

(To list an event, email:


back to the top


– Pre-register:


– Parallel Nsight:


– Parallel Nsight:

CUDA Registered Developer Program

– Sign up:


– List of CUDA-enabled GPUs:

CUDA on the Web

– Learn more about CUDA on CUDA Zone:
– See previous issues of CUDA: Week in Review:
– Follow CUDA & GPU Computing on Twitter:
– Network with other developers:
– Stay tuned to GPGPU news and events:
– Check out the NVIDIA Research page:

CUDA Recommended Reading

– CUDA books:

CUDA Recommended Viewing

– GTC Express:
– SC11 presentations:


CUDA is a parallel computing platform and programming model invented by NVIDIA. It enables dramatic increases in computing performance by harnessing the power of the graphics processing unit (GPU). NVIDIA provides a complete toolkit for programming on the CUDA architecture, supporting standard computing languages such as C, C++ and Fortran. Send comments and suggestions on the newsletter to:
Click here to opt in specifically to CUDA: Week in Review.
Copyright © 2012 NVIDIA Corporation. All rights reserved. 2701 San Tomas Expressway, Santa Clara, CA 95050.