Tuesday, 15 May 2012



Link to NVIDIA

Live: NVIDIA CEO Jen-Hsun Huang Kicks Off GTC

Posted: 15 May 2012 10:17 AM PDT


It's Day One of the GPU Technology Conference (GTC), and that means we're getting ready for NVIDIA CEO Jen-Hsun Huang to hit the keynote stage with some big GPU news.

For those of you unable to attend GTC 2012 in person, we've arranged a snazzy livestream of Jen-Hsun's kickoff keynote presentation. You can find the GTC keynote livestream here.

And, to make sure you get blow-by-blow coverage of Jen-Hsun's big GPU announcement, check out the keynote liveblog below.

GTC Poster Session Shows Breadth of GPU Research

Posted: 14 May 2012 09:17 PM PDT


The GPU computing poster board brigade has arrived at the San Jose McEnery Convention Center, home of NVIDIA's GPU Technology Conference (GTC).

GPUs are typically associated with gaming PCs, mobile phones and video game consoles. But GPU technology has come a long way, and its impact is becoming much more widespread.

For proof, you need only wander through the concourse's display of research poster boards, which give attendees a glimpse at the diversity of research being conducted using GPU technology.

No shortage of poster boards here

Evgeny Salnikov, a research engineer at Microsoft, stopped to admire one display detailing the work being done at the San Diego Supercomputer Center at University of California at San Diego. A team there is using GPUs to accelerate molecular dynamics and assist in the development of transformative drugs. While the findings don't have much relevance to what his work, he said he would do more research on the topic after the conference – he'd never heard of GPUs being used in molecular dynamics.

That's just the tip of the iceberg.

The posse of posters highlights how GPU technology is making possible what would have been inconceivable just a few short years ago. And, in keeping with the increasingly international flavor of GTC, this work is happening all over the world. Here are a few examples:

  •  University of Sheffield's Department of Automatic System Control and Engineering is sharing how it's been using GPU clusters to simulate large-scale pedestrian evacuations. The goal is to help emergency response officials adjust evacuation plans in real time as conditions change.
  • A team from the Graduate School of Informatics at Middle East Technical University in Ankara, Turkey, has been using NVIDIA's CUDA parallel computing platform and programming model to speed up camera sabotage detection. They believe this will reduce lapses in security and preserve computational power for other tasks such as object tracking and activity analysis.
  • Researchers from The Mitre Corp. in McLean, Va., a nonprofit that assists federal agencies by applying systems engineering and advanced technology to national problems, are reporting on efforts to use GPU technology to speed up fingerprint matching algorithms and allow quicker matching of fingerprints against databases containing tens of millions of identities.

Most of the posters on display generate theoretical discussions of possible applications, but some generate more practical interest. One such piece of research came from the Centro de Investigación en Matemáticas in Guanajuato, Mexico. Researchers have been tapping GPU technology to develop what is essentially a virtual green screen for video editing. The goal is to make it possible to superimpose part of a video over a particular part of another image or video, accounting for shadows, people walking in front of the resulting imagery and the like, and making all the necessary adjustments in real time.

These capabilities were of special interest to Dihong Tian, a technical leader at Cisco Systems who works on video conferencing systems. Tian says such capabilities have been prevented by computing and performance limitations, but that he'd be looking further into the Centro de Investigación en Matemáticas' findings to determine if it's feasible to implement them into future products.

Want more info on the poster boards being displayed at GTC? Find all GTC poster board abstracts here. (Use these filter settings: Event = GTC, Year = 2012, Format = Poster)

GPU Tech Conference Off to a Fast Start in San Jose

Posted: 14 May 2012 07:24 PM PDT

GTC 2012 - SJCC

NVIDIA's GPU Technology Conference is back, bigger and better than ever.

Many of the several thousand GTC 2012 attendees began pouring in to the San Jose McEnery Convention Center early Monday morning to take advantage of a full day of pre-conference sessions, and start networking and sharing ideas with their peers. Over the days ahead, they'll fill downtown San Jose's hotels and restaurants with GPU-related buzz.

Early attendees are sitting in on round-table discussions on multi-display systems and GPU-based video processing, or attending GTC 2012 lecture sessions on topics from debugging GPU programs to optimizing GPU algorithms.

But the pace really kicks in on Tuesday, when NVIDIA CEO and co-founder Jen-Hsun Huang delivers big news during his highly anticipated keynote address. For those unable to attend GTC 2012 in person, we'll provide a livestream of Jen-Hsun’s GTC keynote here.

The GTC 2012 agenda wall makes it easy to
navigate the hundreds of educational sessions
throughout the show

Attendees will then choose from hundreds of sessions over the ensuing days, each with details of how GPU technology is advancing various industries and powering cutting-edge research.

For instance, Oak Ridge National Laboratory will share how they've been using GPUs to build Titan – the world's fastest supercomputer. Beijing's BGI will talk about their use of GPUs for bioinformatics research. Attendees can get the scoop on the latest advances in stereoscopic 3D visualization from the Tokyo Institute of Technology, winner of last year's Gordon Bell Prize. And, the Emerging Companies Summit will offer attendees an inside look at how startups are using GPUs.

In Wednesday's keynote, Princeton's Iain Couzin will discuss his use of GPU technology to study collective behavior in the natural world. Couzin's research explores how individual behavior can produce dynamic collective behaviors, and shed light on how our brains work and what drives crowd behavior. It also highlights how GPU technology can have an unexpected impact.

Finally, after two days of sessions and conversations with peers, attendees will be ready to blow off some steam at Wednesday night's GTC 2012 party, featuring a live performance that will test the laws of physics.

With world-class speakers and plenty of opportunities for attendees to share stories and experiences, GTC 2012 is the year's signature GPU technology event. And it's not too late to attend. For more information, visit the GTC 2012 conference registration page.

Oh, and check out the pre-conference photos below.

GTC 2012 - agenda wall GTC 2012 - SJCC gtc-banner-san-jose-downtown gtc-preconference-registration gtc-pre-conference-session gtc-c++-lounge gtc-preconference-session-1` Taxi_Topper_4 SJ_Airport_Pictures 008 IMG_2875


U.S. Energy Dept: Supercomputing Changes Energy Game

Posted: 14 May 2012 11:36 AM PDT


U.S. Energy Secretary Steve Chu recently described to Forbes how supercomputing is bringing change to our energy future, as well as the overall pursuit of science.

We couldn't agree more.

The U.S. Department of Energy (DOE) is backing up this belief by investing in systems such as the GPU-accelerated Titan supercomputer at the Oak Ridge National Laboratory (ORNL) in Tennessee. Titan will be one of the fastest machines in the world when it is deployed.

Scientists at DOE laboratories are using supercomputers to simulate everything from how viruses attack biological cells to improving the fuel efficiency of combustion engines, and even developing cleaner, more sustainable energy source alternatives.

Powerful GPU-based supercomputers, like Titan, are key to performing computer simulations, which have become the third pillar of science – joining theory and experimentation as a means of discovering new phenomena or testing hypotheses.

The Oak Ridge National Laboratory is
home to the Titan supercomputer

Simulating physical and biological systems is often much easier than performing experiments in a laboratory. For example, determining the exact temperature required to make a chemical reaction happen in a lab is not only labor intensive, but also time consuming. A software program running on a supercomputer can easily model the same chemical reaction at precise temperatures and other environmental conditions.

The challenge with this kind of research is that accurate computer simulations require enormous computing performance from systems consisting of tens of thousands of computer servers – and they simply require too much power to run. The current Jaguar supercomputer at ORNL, for example, delivers about 1 petaflop of computing performance while consuming 7 megawatts of power – equivalent to the power requirements of a small town!

This is why ORNL is building Titan using traditional x86 CPUs accelerated by NVIDIA Tesla GPUs. GPU-accelerated computing requires much less power than CPU-only supercomputers.

The Tokyo Institute of Technology (popularly called Tokyo Tech) made a similar choice. They recently deployed the Tsubame 2.0 GPU-accelerated supercomputer as the greenest petaflop system in the world. Tsubame 2.0 consists of about 1,500 computer servers (about 40 racks) accelerated by the Fermi-based Tesla M2050 GPUs from NVIDIA. It can deliver 1 petaflop of sustained performance while consuming 1.3 megawatts. In contrast, it would take 4,000 non-accelerated, CPU-only servers (about 100 racks) to build a similar petaflop supercomputer, and it would consume between 2 to 3 megawatts.

Scientists and engineers will always desire greater computing capability in their quest to advance the frontiers of science. The DOE's commitment to providing them the best, most energy-efficient supercomputing tools will help the U.S. continue to lead the world in technology and engineering.

No comments:

Post a Comment