Communications of the ACM
Innovation Promises to Cut Massive Power Use at Big Data Companies in a Flash
Vivek Pai, an associate professor of computer science at Princeton, worked with graduate student Anirudh Badam to develop a software technique that could radically cut power use.
Credit: Frank Wojciechowski
Princeton University researchers have developed SSDAlloc, software that could help organizations use a type of memory in their servers that requires much less energy than conventional systems.
SSDAlloc enables organizations to substitute flash memory for the more expensive and energy-intensive random access memory (RAM) that is currently used for most computer operations.
"The biggest potential users are the big data centers," says Princeton professor Vivek Pai.
He says flash uses much less electricity than RAM, so switching memory types can reduce a company's power bill by up to 90 percent. SSDAlloc changes the way that programs look for data in a computer, according to Princeton graduate student Anirudh Badam. Flash memory is much faster than a hard drive, but that led to bottlenecking, which has limited its potential applications. The researchers say SSDAlloc moves flash memory up in the internal hierarchy of computer data, and tells the computer to consider it a larger, slower version of RAM. "Our system monitors what the host system is doing and moves it into and out of RAM automatically," Pai says.
The researchers say SSDAlloc can be applied to a whole class of applications for use in many different fields.
From Princeton University
View Full Article
Abstracts Copyright © 2012 Information Inc., Bethesda, Maryland, USA
No entries found