Communications of the ACM
Researchers Save Electricity With Low-Power Processors and Flash Memory
Each of the FAWN cluster's 21 nodes has a low-power off-the-shelf processor and a 4-Gbyte compact flash card.
Credit: Carnegie Mellon University
During ACM's recent Symposium on Operating Systems Principles, researchers from Carnegie Mellon University (CMU) and Intel Labs Pittsburgh (ILP) won the best paper award for their article on Fast Array of Wimpy Nodes (FAWN), a server architecture that can handle data-heavy applications with greater speed and efficiency than current systems.
To create FAWN, CMU computer scientist David Anderson and colleagues collaborated with ILP to integrate netbook processors with flash memory. Flash memory works more quickly than hard disks, costs less than DRAM chips, and is the most energy efficient option available. To test the server architecture, the researchers constructed a FAWN computing cluster out of 21 nodes, each equipped with a low-cost, off-the-shelf processor and a four-gigabyte flash card. They discovered that the FAWN cluster could manage 10 to 100 times more requests than a disk-based one while using the same amount of energy.
Researchers are now constructing a FAWN cluster that uses Intel's Atom processor. "FAWN systems can't replace all of the servers in a data center, but they work really well for key-value storage systems, which need to access relatively small bits of information quickly," Anderson says. The researchers hope to use FAWN for data analysis in the future, he says.
From Carnegie Mellon News
View Full Article
Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA 
No entries found