U.S. research lab focuses on building computers of the future
- 31 May, 2014 06:15
Computer scientists at Sandia National Laboratories have launched an effort to develop whole new types of computers that will be used 10, 25 or even 50 years from now.
Researchers at Sandia are developing next-generation supercomputers, nano-based computing, quantum computing and brain-inspired computing. Sandia, in Albuquerque, N.M., is a contractor for the U.S. Department of Energy and focuses its work on technology that supports national security.
"We think that by combining capabilities in microelectronics and computer architecture, Sandia can help initiate the jump to the next technology curve sooner and with less risk," said Rob Leland, head of Sandia's Computing Research Center.
Erik DeBenedictis, a research scientist at Sandia, said it's critical to focus on what might be the next generation of computers because traditional computers won't be able to keep up with Moore's Law much longer. (Moore's law was posited by Intel co-founder Gordon Moore, who predicted in 1965 that that the number of transistors on a chip would double about every two years and could be done inexpensively.
For years, industry analysts and insiders have claimed that the death of Moore's Law is imminent. So far, the industry has been able to keep up with it. DeBenedictis said he believes that time is running out.
"It's absolutely clear that the reduction in the size of wires is going to end pretty soon," he told Computerworld. "You can't have wires that are smaller than an atom. And now they're getting small enough that they can't conduct electricity very well. A reason the law is slowing down is that we can make transistors smaller but their energy consumption doesn't go down as fast as it used to."
That means there will continue to be advances in traditional computing, but they won't be as fast-paced or revolutionary as they have been.
To continue the pace of advancements, the computing industry will have to turn at least some of its attention to new computer platforms.
"In order to get the kind of performance we need to take on new problems in science, not to mention to drive the massive amounts of data that we're all generating and using in our everyday lives, we'll need to have new kinds of technology that are much more efficient and that can be eventually manufactured at an affordable cost," said Dan Olds, an analyst with The Gabriel Consulting Group.
He added that we're not at the end of the line with Moore's Law, but he can see ahead to where it will end if the industry continues to incrementally shrink and then speed up frequencies in CPUs.
Because of that, now is the time to push research into new types of computers.
"Yes, we're definitely at a point where we could use some breakthroughs," said Olds. "I believe that we'll reach the point of diminishing returns from 'shrinks' at about the 9nm to 10nm process level. This means in the absence of a technological breakthrough, we'll need to improve other parts of the computing process to keep us on the same price/performance curve as Moore's Law."
One of the new forms of computing that scientists at Sandia are working on is a system that works more like a brain than a conventional computer.
A brain-inspired computer is expected to be able to tackle real-world situations in real-time, while also being able to run on the same power as a 20-watt light bulb.
The only "machine" that can handle those functions is the human brain. Sandia researchers are hoping to change that.
"Today's computers are wonderful at bookkeeping and solving scientific problems often described by partial differential equations, but they're horrible at just using common sense, seeing new patterns, dealing with ambiguity and making smart decisions," said John Wagner, cognitive sciences manager at Sandia, in a recent statement.
Sandia also is working on quantum computing, which has become something of a lightning rod in the high-tech industry.
Quantum computing, which is based on the tenants of quantum physics, is not only the holy grail of supercomputing, but it's also so complicated that it's tough for even computer scientists, as well as physicists, to understand.
Many believe these computers will be able to surpass the top classic supercomputers in some calculations, especially problems that require searching through huge amounts of data, or finding answers to questions so complex that machines like IBM's Blue Gene and Cray's supercomputing systems might need hundreds of years to solve them, or might never solve them at all.
While many in the industry say it will take as long 50 years to seeing a real quantum machine, D-Wave, a quantum computing company based in Burnaby, British Columbia, said it is building them. D-Wave has sold the computers to Google and NASA.
DeBenedictis declined to get into the state of quantum computing today but said it's important to continue research into new areas of computing to handle the changing and growing ecosystem of applications that will be used.
"The computers we have today are good at numerical calculations," he said. "But think about how we use them today. We use Google and social media and we watch videos and process text. There's an evolution happening in the applications we need our computers to handle. I would say [future computer platforms] will be better suited to the way we want to use computers tomorrow."
This article, U.S. research lab focuses on building computers of the future, was originally published at Computerworld.com.
Sharon Gaudin covers the Internet and Web 2.0, emerging technologies, and desktop and laptop chips for Computerworld. Follow Sharon on Twitter at @sgaudin, on Google+ or subscribe to Sharon's RSS feed. Her email address is email@example.com.
Read more about high performance computing in Computerworld's High Performance Computing Topic Center.