Carnegie Mellon University (CMU) researchers have developed GraphChi, a framework for running large-scale computations on one personal computer (PC).
The researchers say the software could help developers create a recommendation engine using social network connections.
GraphChi takes advantage of the ample hard disk space available in modern PCs to complete graph computations. For example, CMU's Carlos Guestrin says a Mac Mini running GraphChi can analyze Twitter's social graph from 2010, which contains 40 million users and 1.2 billion connections, in 59 minutes.
"The previous published result on this problem took 400 minutes using a cluster of about 1,000 computers," Guestrin notes.
Graph analysis also drives the development of new Web products, says the Massachusetts Institute of Technology's Jeremy Kepner. "Enabling Web developers to construct these analyses on their desktop computers catalyzes these industries and accelerates product development," Kepner says.
In addition, GraphChi can analyze streaming graphs, which more accurately simulate large networks by showing how relationships change over time. "Tools like GraphChi will let many companies and startups solve all their graph-computing needs on a single machine," Guestrin says.
From Technology Review
View Full Article
Abstracts Copyright © 2012 Information Inc., Bethesda, Maryland, USA