U.S. HPC lead in danger

The DOE's exascale program is stalled, and its prospects appear bleak at a time of budget-cutting fever.

The SC11 supercomputing conference in Seattle last month saw an almost obsessive focus on the development of an exascale computing system -- one that would be roughly 1,000 times more powerful than any existing system -- before the end of the decade.

Programs to develop ever-higher-performance computing systems are underway around the world -- but the U.S. isn't yet leading the way, even though it has dominated HPC development for decades.

China and Europe, in particular, are moving ahead with strong exascale programs. And Japan is increasingly picking up the pace. Efforts in the U.S., however, are stalled; the federal government hasn't released a formal exascale plan since last summer, when it outlined preliminary goals of building an exascale system that doesn't use more than 20 megawatts of power by 2019 or 2020.

The U.S. Department of Energy is due to deliver to Congress by Feb. 10 a timetable and budget for building an exascale system.

The delivery of the U.S. plan couldn't come at a worse time politically, particularly after the Congressional Super Committee last month failed to reach a budget agreement, triggering significant mandated cuts.

Experts say exascale systems could be capable of solving the world's greatest scientific problems. If the U.S. falls behind, the research would increasingly be done in other countries.

Exascale development efforts could also seed new processor, storage and networking technologies. Breakthroughs in other countries could give rise to new challenges to U.S. tech dominance.

Many U.S. scientists have warned of the problems posed by the strong exascale projects underway in Europe and China.

"The EU effort is more organized at this stage ... with strong backing from the European Commission," said Jack Dongarra, a computer science professor at the University of Tennessee. "The bottom line is that the U.S. appears stalled and the EU, China and Japan are gearing up for the next generation."

Corporate interest in high-performance computing is growing quickly, and many IT operations will likely continue to seek more powerful HPC systems, whether created in the U.S. or not.

Even midsize companies like L&L Products are using supercomputers to create new products and boost sales.

L&L, a maker of structural composite products for the automotive industry, used the technology to create virtual models and run automotive crash simulations, among other things.

Steven Reagan, the computational modeling manager at L&L, said the 53-year-old company has doubled its business since first adopting HPC technology six years ago.

If the U.S. does fall behind, observers wonder whether it can get involved in the next computing level -- zetascale.

If the current pace of HPC development continues, a zetascale system can be expected around 2030. But heavy research is required: No one knows what it would look like, or whether it's even possible without using entirely new approaches like quantum computing.

This version of this story was originally published in Computerworld's print edition. It was adapted from an article that appeared earlier on Computerworld.com.

Copyright © 2011 IDG Communications, Inc.

It’s time to break the ChatGPT habit
Shop Tech Products at Amazon