Silicon chips with errors could help boost computer power

Subscribe to Oneindia News

London, May 26 (ANI): An Indian researcher has said that to ensure perfect operation of silicon chips, manufacturers are running them at a higher power than actually needed.

Researchers are saying now that silicon chips that have errors in them could actually help in boosting the performance of computers.

As components shrink, chipmakers struggle to get more performance out of them while meeting power needs.

The silicon industry is defined by Moore's Law, which predicts that the number of transistors that can fit on a given area of silicon, for a given price will double every 18-24 months.

For Professor Rakesh Kumar at University of Illinois, the demise of Moore's Law is being hastened by an insistence on making silicon chips operate flawlessly.

Professor Kumar said variations in manufacturing, environment, and workload can conspire to make a chip suffer errors. Manufacturers try to ensure that whatever happens, he said, the chip works correctly.

"It's a case of 'if the software asks the chip to do something it does it at any cost,'" BBC News quoted him as saying.The researcher said that as components shrink further and further, flawless execution is becoming harder and errors, inevitable.

And so, he advocates "stochastic processors" - which are subject to random errors.

"The hardware is already stochastic so why continue pretending it's flawless?" he asked. "Why put in more and more money to make it look flawless?"

In many cases the errors will not have a significant impact on the workings of a computer. In other cases, he said, they could cause a system to crash and so, it is up to the designer to decide how much error should be tolerated.

Professor Kumar and colleagues are investigating alternatives to cope with this issue - re-writing it so an error simply causes the execution of instructions to take longer or logging a user's actions, and then using the log to spot when something unexpected occurs. (ANI)

Please Wait while comments are loading...