RICE (US) — Trimming the fat from rarely used portions of a computer chip can make it twice as fast, able to consume half the energy, and take up half the space as traditional microchips.
“I believe this is the first time someone has taken an integrated circuit and said, ‘Let’s get rid of the part that we don’t need,'” says Krishna Palem, professor of computing at Rice University.
“What we’ve shown is that we can boost performance and cut energy use simultaneously if we prune the unnecessary portions of the digital application-specific integrated circuits that are typically used in hearing aids, cameras, and other multimedia devices.”
Pruning is the latest example of “inexact hardware,” the key approach that the Institute for Sustainable and Applied Infodynamics (ISAID) is exploring with Switzerland’s Center for Electronics and Microtechnology (CSEM) to produce the next generation of energy-stingy microchips.
The concept is deceptively simple: Slash power demands on microprocessors by allowing them to make mistakes. Managing the probability of errors and limiting which calculations produce errors, cuts energy demands while boosting performance.
“Our initial tests indicate that the pruned circuits will be at least two times faster, consume about half the energy and take up about half the space of the traditional circuits,” says graduate students Avinash Lingamneni. It’s hoped the system will perform even better in the final tests, still under way.
“The cost for these gains is an 8 percent error magnitude, and to put that into context, we know that many perceptive types of tasks found in vision or hearing applications can easily tolerate error magnitudes of up to 10 percent,” says Christian Enz, who leads the CSEM arm of the collaboration and is a study co-author.
The next hurdle for “pruning” will be to use the technique to create a complete prototype chip for a specific application. Design on a chip for a hearing aid is expected to begin this summer, Palem says.
“Based on what we already know, we believe probabilistic computing can produce application-specific integrated circuits for hearing aids that can run four to five times longer on a set of batteries than current hearing aids.”
More news from Rice University: www.media.rice.edu/media/