(Santa Clara, California--March 19, 2013) KICKING THINGS OFF — TITAN IN HIGH GEAR — DIGESTING PI WITH GPUs — PYTHON TAPS GPUs — SWEATING BULLETS — ABOUT THAT AFTERNOON JOLT
TODAY’S HOTTEST TICKETS
Keynote Kick Off. Get up early. Gulp down your coffee. Grab a good seat for the Day One opening keynote. NVIDIA’s Jen-Hsun Huang gives a great show. Expect a sweeping view of what’s next in computing and graphics, replete with on-stage guests. 9am, Hall 2.
Architecture Goes Digital. Learn how one of the world’s top architectural firms uses digital simulation and analysis – on the desktop, across the WAN, in the cloud – before the first shovel hits the ground. Ken Sanders, a managing director at Gensler, shares case studies from the firm’s global portfolio, including new HQs for PNC and NVIDIA, as well as Shanghai Tower – soon to be China’s tallest building. 1pm, Room 111.
First ARM + GPU Cluster for HPC. See a developed first proof of concept for a hybrid platform using an ARM multicore CPU for energy efficiency, and a discrete GPU accelerator for compute performance. Barcelona Supercomputing Center’s Alex Ramirez discusses the system’s architecture and software stack, preliminary performance and power measurements, and guidelines for future such platforms. 1:30pm, Room 210E.
Titan Hits High Gear. Early outcomes from Titan, the world’s fastest supercomputer, feature in this talk by Fernanda Foertter and Jack Wells of Oak Ridge National Laboratory. They’ll cover preliminary early science results from users – as well as details about Titan’s system setup and how to apply for time on its 18,688 GPU-accelerated nodes. 2pm, Room 210E.
Champion Pi Eater Tells All. Ed Karrels, of Santa Clara University, reveals how he computed pi to the four-quadrillionth bit, stealing the record from Yahoo! by wielding a single CUDA-enabled computer. 2pm, Room 230B.
4Cs Come to GTC. The diamond’s dance of light has entranced admirers for ages. Its beauty happens not by chance but by the design and craft of its cut. NVIDIA’s Rupali Deshpande talks about how GPUs accelerate converting rough stones to sparkling gems. 2:30pm, Room 230B.
Steering Sailing Into a Spectator Sport. The America’s Cup’s 72-foot catamarans, sailing at over 40 knots, will be a spectacular sight when the venerable race takes place in San Francisco this summer. Serious Intent’s Tim Heidmann explains how GPUs help perform key tasks in the broadcast pipeline, so broadcasters can better convey the excitement to an international TV audience. 5:30pm, Room 210B
LATE BREAKING NEWS
S3200 — Estimating Fracture Size in Gas Shale Formation Using Smart Proppants and GPGPU, scheduled for Tuesday at 1400, in Room 212A is now in Room 112, same time.
S3556A — System Design of Kepler Based HPC Solutions, scheduled for Tuesday at 1400, in Room 113, is now in Room 111, same time.
S3504 — RDMA for GPUDirect - Interconnect Optimizations for High Performance Computing, scheduled for Tuesday at 1500, in Room 113, is now in Room 111, same time.
S3551 — Real-time Lighting and Rendering for Architectural Visualization, scheduled for Tuesday at 1400, in Room 111, is now in Room 113, same time.
S3569 — Mobile Augmented Reality in Manufacturing. What Can It Do For You? scheduled for Tuesday at 1430, in Room 111, is now in Room 113, same time.
S3052 — Deploying General Purpose GPUs in a Manufacturing Environment: Software and CFD Development in Bicycle Manufacturing, scheduled for Tuesday at 1500, in Room 111, is now in Room 113, same time.
S3431 — Capturing, Processing, and Streaming Immersive Video in Realtime, scheduled for Tuesday at 1500, in Room 210B, is cancelled.
WHAT YOU MAY HAVE MISSED
The GTC poster reception spilling across the main lobby has 150+ works from researchers as near as Stanford University and as far afield as Saudi Arabia’s King Abdullah University of Science and Technology, and Cuba’s Advanced Technologies Application Centre. They’re here for the duration of the show, so stop by and chat up authors on their latest projects.
Python programming language now supports GPUs. Yesterday’s first GTC announcement means that the growing ranks of programmers using the Python open-source language can take full advantage of GPU acceleration for HPC and big data analytics applications.
Penguin Computing’s session will discuss the workflow for using Reality Server on POD. Using Fluid Inc.’s Configure offering as an example, the session will illustrate how retailers can leverage POD, NVIDIA Tesla GPUs and the Reality Server platform for scalable, fast-to-market and easy-to-manage product customizations.
HEARD ON THE FLOOR
Sweating bullets. Fazel Adabi had clearance of only an inch or so on either side while sliding the new Lamborghini Aventador into the Convention Center’s narrow doorway. That’s tough with a car not built for nudging forward gently.
Afternoon jolt. Conspiracy theorists were only slightly assuaged to learn what caused the false fire alarm that cleared out the Convention Center around 2pm. The bells, lights and looped announcements were set off by earth-shaking construction work, which is now stopped for the rest of the show.
Lean In. Facebook COO Sheryl Sandberg may be hitting best-seller lists in the New York Times, but the top seller so far at GTC is David Kirk and Wen-mei Hwu’s Programming Massively Parallel Processors.
Exotic accents. A table of grizzled GTC veterans at the San Jose Convention Center for the third-straight year were heard admiring, at length, the teak wood accents now installed as part of the makeover.
WHAT THEY’RE TWEETING
David K. Johnson @david_kjohnson My goal for the next 2 days is to understand precisely how firms are applying GPU-accelerated #VDI to advanced engineering uses. #gtc13
David K. Johnson @david_kjohnson At the GPU Technology Conference. The attendees here are the real deal. Physicists, geologists, engineers. People with a job to do. #gtc13
Sunjay Dhama @sdsunjay Accelerate your life, learning, and passion. #GTC2013
Dan Rosanova @DanRosanova #GTC2013 is turning out to be an amazing event and is packed! Some serious tech here. Awesome event for a true CS techie.
Acceleware Ltd. @acceleware Packed Acceleware tutorial at @nvidia #GTC13 even after being moved to a larger room due to high demand: http://ow.ly/i/1I4YM
Patrick Moorhead @PatrickMoorhead Headed to #GTC13 in Santa Clara.
Kenneth O’Brien @ken_obrien Must submit a poster for #GTC14. Too late for #GTC13. Looks like a lot of cool stuff there.
HPC Guru @HPC_Guru Barcelona Supercomputing Center to discuss Pedraforca #ARM+#GPU Cluster for #HPC, Mar. 19 at #GTC13 http://ow.ly/gQcHO