Friday 29 October 2010

nTersect

nTersect


Tesla GPUs Power #1 Supercomputer

Posted: 28 Oct 2010 12:34 PM PDT

This week was a momentous one in the world of supercomputing.

Today in Beijing, the Tianjin Supercomputing Center revealed details of Tianhe-1A system – a new GPU-enabled supercomputer that delivers 2.5 petaflops of performance – that's 2,500 trillion (2,500,000,000,000,000) floating point operations per second  - the equivalent of 175,000 laptops!

If you tried to build a system capable of this level of performance using just CPUs, you'd need more than 50,000 of them, consuming 12 megawatts of power. Instead, the system uses a "heterogeneous" model – massively parallel GPUs and multi-core CPUs – to deliver more performance, more efficiently. The system uses both Tesla GPUs along with CPUs and, thanks to this design, Tianhe-1A consumes only 4 megawatts, making it 3 times more power efficient than a CPU-only machine. The power saved by using GPUs is enough to power your home for 8000 years!  Give or take.

New Image

Many supercomputing centers are now building GPUs into the systems to get big boosts in performance, while staying within tight power constraints. Already on the Top500 list are systems from Tokyo Institute of Technology, Chinese Academy of Sciences, and they are engaged in some fascinating areas of research such as tsunami modeling and particle simulation……and there are more coming.

The massively parallel architecture of GPUs, coming from its graphics heritage, is now delivering transformative results for scientists and researchers all over the world. For some of the world's most challenging problems in medical research, drug discovery, weather modeling, and seismic exploration – computation is the ultimate tool. Without it, research would still be confined to trial and error-based physical experiments and observation. 

With GPUs, supercomputing is no longer just for privileged, well-funded organizations. This mass availability of processing power is starting what we like to think of as a race for better science, as more countries develop this capability and begin to search for solutions to some of life's greatest challenges.

NVIDIA recently posted a pilot for a new documentary series entitled "The 3rd Pillar of Science" – watch the trailer below, or see the full episode here. This pilot investigates the role of computation in medicine, and speaks to medical experts in fields such as cancer treatment and heart surgery to see how they are using GPUs to transform their work.  We hope you enjoy it and would love to hear what you think about it in the comments section.

No comments:

Post a Comment