Nvidia eyes an exascale supercomputing paradigm

Posted by TG Daily Staff

Simulations powered by petaflop-capable HPC systems help scientists and policy-makers improve their understanding of a wide range of topics.

However, there are serious exascale-class problems that simply cannot be solved in any reasonable amount of time with the computers we have today.



Nvidia outlines the future of exascale supercomputingIndeed, future exaflop systems are expected to be 1,000 times faster than petaflop computers - delivering a staggering one billion billion calculations per second.


But power, or lack thereof, is hampering the adoption of a realistic exascale computing paradigm.



"An exascale computer using today's x86 technology would require two gigawatts of power, equivalent to the maximum output of the Hoover Dam. Our technology needs to be about 100 times more energy efficient in order to build practical exascale systems," explained Steve Scott, CTO of Nvidia's Tesla group.

"[Unfortunately], we can no longer keep dropping the chip voltage with each reduction in transistor size. The result is that power has become the dominant constraint in processor design. If we ran all the transistors we could put on a chip at full speed, the chip would melt."



That's where GPUs come in, says Scott. Because unlike traditional CPUs, which are designed to make serial tasks run as quickly as possible, GPUs execute multiple parallel tasks in a power-efficient manner. 



As Scott notes, the next-gen Kepler GPUs used in the recently announced Titan system will provide more than one teraflop of performance per chip. 



"In heterogeneous computing, the GPU can perform the heavy lifting, executing the parallel work with very low power, and the CPU can then quickly execute the serial work that's left over... This is the only hope of getting to exascale computing at reasonable cost and power."

However, Scott acknowledged that adopting an exascale HPC model certainly won't be an "easy" endeavor.

"Imagine telling the auto industry you need to develop a car that goes 1,000 times faster and is 100 times more energy efficient.

"Yes, that's [definitely] a very tall order. But I'm confident heterogeneous solutions with GPUs are the right path to get us there," he added.