Share:   Facebook Twitter Linked-in Google Reddit Stumbleupon email
CUDA Week in Review Newsletter
Tue., June 18, Issue #95 Newsletter Home


Welcome to CUDA: WEEK IN REVIEW, a news summary for the worldwide CUDA, GPGPU and parallel programming community.
CUDA Pro Tip: The new Understand Fat Binaries and JIT Caching Parallel Forall blog post offers an overview of CUDA "fat binaries" and compilation for multiple GPU architectures, as well as just-in-time PTX compilation for forward compatibility.


GPU-Accelerated Guidance and Control for Robotic Systems
This week’s Spotlight is on Jon Rogers of the Department of Aerospace Engineering at Texas A&M University. Jon is director of the Helicopter and Unmanned Systems Laboratory, where he works on new technologies for autonomous systems.

Read our interview with Jon Rogers.
Jon Rogers


CUDA 5.5 RC Includes ARM Support
NVIDIA announced that the CUDA 5.5 Release Candidate is now available to the public. Features include native compilation on ARM-based platforms; enhanced Hyper–Q support; MPI workload prioritization; and new guided performance analysis.
Download CUDA 5.5 RCCUDA 5.5 webinar (June 25) | Become a registered developer

Barcelona Taps Tesla, Tegra for Next-Gen Hybrid Supercomputer
The Register reports that Barcelona Supercomputing Center will combine "a baby ARM processor aimed at smartphones and tablets with a full-on GPU coprocessor… With the Kayla system board, which comes in a Mini-ITX form factor, the CPU side is again a Tegra 3 chip from NVIDIA, but the board has a PCI Express 2.0 x16 link to hook a full-on Tesla GPU to the ARM processor."

Titan Completes Acceptance Testing
Oak Ridge National Lab's Titan supercomputer has completed acceptance testing to ensure the functionality, performance and stability of one of the world's most powerful supercomputing systems for open science. It is the first to combine different types of processing units to maximize performance at such a large scale.

New CUDA Book
Congrats to author Nicholas Wilt on his new book titled "The CUDA Handbook: A Comprehensive Guide to GPU Programming." The book covers hardware and software, and provides more than 25,000 lines of open source code for developers. The book is available for purchase here. The publishers are offering a 35% discount code for CUDA newsletter readers: CUDAHBNVIDIA.

NMath Premium
CenterSpace Software released NMath Premium, an easy–to–use library that automatically detects the presence of a CUDA-enabled GPU at runtime and seamlessly redirects appropriate computations to it. The library can be configured to specify which problems should be solved by the GPU, and which by the CPU. If a GPU is not present at runtime, the computation automatically falls back to the CPU without error.

INCITE Call for Proposals (Ends June 28)
Accelerate your research on Titan, the world’s #1 supercomputer for science, by harnessing more than 20 petaflops of parallel processing using GPUs. Find out more.

Upcoming Webinars
June 20: GPU-Accelerated XenDesktop, Thomas Poppelgaard,
June 25: CUDA 5.5 Features and Overview, Ujval Kapasi, NVIDIA
June 26: Understanding Dynamic Parallelism with Allinea’s Unified Tools, Mark O’Connor, Allinea
July 9: NVIDIA GRID VCA: A Turnkey Appliance Delivering Remote Graphics, Ankit Patel, NVIDIA
July 10: Intro to the CUDA Toolkit as an Application Build Tool, Adam DeConinck, NVIDIA


back to the top
Need a CUDA trainer? See list of worldwide CUDA trainers and consultants.
Want to get smarter? Enroll in the open online course Intro to Parallel Programming.
Have CUDA questions? Check out NVIDIA DevTalk forums and Stack Overflow.
Require fast access to docs? Visit the CUDA doc library.


back to the top
Have you been to a Meetup yet? Join one or start your own. Upcoming events:
Paris GPU Meetup, June 20
Rome GPU Meetup, June 21
New York GPU Meetup, June 24
Brisbane GPU Meetup, June 27
Toronto GPU Meetup, June 27


back to the top
Parallel Forall:
Understand Fat Binaries and JIT Caching, by Mark Harris
Clean Up After Yourself to Ensure Correct Profiling, by Mark Harris

GPU at Heart of World’s Fastest Supercomputer Hits Laptops, by Mark Ebersole


back to the top
Pelican Imaging seeks a Platform Software Architect/Principal Engineer to work on next generation embedded software running on multiple hardware families and OSs (Android, IOS, Windows...). Requires ability to work from high level software architecture development to detailed optimization on various processors and GPU families.


back to the top


Apple Final Cut Pro User Group (SF Cutters)
  June 19, 2013, 6:30 pm, San Francisco, Calif.
Speakers: Bob Sarles, Mark Christiansen and NVIDIA

GPU Accelerated XenDesktop for Designers and Engineers (GTC Express Webinar)
  June 20, 2013
Presenter: Thomas Poppelgaard,

4-Day CUDA Course(Acceleware)
  June 25-28, 2013, Calgary, AB, Canada
Instructor: Dr. Kelly Goss, Acceleware

Dynamic Parallelism with Allinea's Unified Tools(GTC Express Webinar)
  June 26, 2013
Presenter: Join Mark O’Connor, Allinea


HPC@LR Sumer School (University of Montpellier 2)
  July 3-5, 2013, Montpellier, France

PUMPS Summer School
  July 8-12, 2013, Barcelona, Spain
Co-Directors: Mateo Valero (BSC, UPC) and Wen-mei Hwu (University of Illinois)

GPU Summer School (University of Alicante)
  July 22-24, 2013 Alicante, Spain

4-Day CUDA Course (Acceleware)
  July 29-August 1, 2013, San Jose, Calif
Instructor: Dr. Kelly Goss

GPU Technology Conference Japan
  July 30, 2013, Tokyo, Japan

(To list an event, email:


back to the top

Webinar Replay

GROMACS and Kepler GPUs, by Erik Lindahl, Stockholm University (1:16:47)

GPU-Accelerated Apps

List of 200+ popular GPU-accelerated scientific and research applications (PDF 402KB).

GPU Test Drive

Want to try Tesla K20 for free? Sign up here.

CUDA Documentation

The new CUDA documentation site includes release notes, programming guides, manuals and code samples.

CUDA Online Courses

Udacity Course
Coursera Course

NVIDIA Developer Forums

Join us on the NVIDIA DevTalk forums to share your experience and learn from other developers. You can also ask questions on Stack Overflow, using the ‘cuda’ tag.

CUDA Consulting

Training, programming and project development services are available from CUDA consultants around the world. To be considered for inclusion on list, email: (with CUDA Consulting in subject line).

GPU Computing on Twitter

For daily updates about GPU computing and parallel programming, follow @gpucomputing on Twitter.



CUDA on the Web

CUDA Spotlights
CUDA Newsletters
GPU Test Drive


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
Copyright © 2013 NVIDIA Corporation. All rights reserved. 2701 San Tomas Expressway, Santa Clara, CA 95050.