Scientists have taken the next step in the evolution of the computer chip, developing self-healing integrated chips.
A team of researchers at the California Institute of Technology (Caltech) said they now can envision smartphone and computer chips not only defending themselves but also repairing themselves, recovering from trouble -- like total transistor failure -- in microseconds.
[ Also on InfoWorld: 5 storage servers battle for business; find out which one prevails in Test Center's NAS shoot-out. | Keep up on the day's tech news headlines with InfoWorld's Today's Headlines: Wrap Up newsletter. ]
The scientists reported that they destroyed different parts of their self-healing chips, which were running power amplifiers, with a high-power laser and then watched the chips automatically develop a work-around to bypass the damage in less than a second.
"It was incredible the first time the system kicked in and healed itself," said Ali Hajimiri, a professor of electrical engineering at Caltech, in a statement. "It felt like we were witnessing the next step in the evolution of integrated circuits. We had literally just blasted half the amplifier and vaporized many of its components, such as transistors, and it was able to recover to nearly its ideal performance."
Self-healing chips in tablets, smartphones and laptops would be a big advance, according to Dan Olds, an analyst with The Gabriel Consulting Group.
"This new technology could help to keep performance at a high level by routing around internal components that have failed or are getting long in the tooth," said Olds. "But programming this capability and implementing it on a large scale could be a pretty complex and expensive undertaking. They've put together an interesting demonstration project, but it's a long way from being placed in any actual device."
He also noted that there are self-healing systems today on high-end servers.
"They can either soldier on through any hardware or software error or very quickly recover from a failure," said Olds. "This is usually done through having the system deconfigure failed components or start using spare parts in the system. It looks to me like the Caltech guys are applying these concepts at the chip level rather than the system level."
According to CalTech, the scientists used numerous on-chip sensors that continually monitor temperature, current, voltage and power. The sensors then feed that information into the central processor, which analyzes the device's performance and calculates if anything needs to be repaired.
However, the processor does not use algorithms that have been told how to respond to every possible problem. It would be difficult, or even impossible, to predict every possible problem. Instead, the processor is built to draw its on conclusions.
"You tell the chip the results you want and let it figure out how to produce those results," said Steven Bowers, a CalTech graduate student and a researcher on the project. "The challenge is that there are more than 100,000 transistors on each chip. We don't know all of the different things that might go wrong, and we don't need to. We have designed the system in a general enough way that it finds the optimum state for all of the actuators in any situation without external intervention."