The NSF last week announced a grant to the Univ. of Illinois, Urbana-Champaign (UIUC), to expand the NSF’s previously announced Cluster Exploratory (CluE). With additional funding and equipment from HP, Intel, Yahoo!, and the UIUC, this expansion gives researchers access to one more cluster for data-intensive computing research.
The UIUC cluster is part of the Cloud Computing Test Bed, one of six “centers of excellence” that the consortium of HP, Intel, and Yahoo! are creating around the world to foster research in data-intensive computing. The NSF has a similar CluE with sponsorship from Google and IBM and academic partners the Univ. of Washington, Carnegie-Mellon Univ., MIT, Stanford, the Univ. of California, Berkeley, and the Univ. of Maryland.
Cloud computing makes use of vast banks of computer servers to make supercomputing available to a large number of users. In the same vein, the TeraGrid integrates high-performance computers and data resources to maximize the high-end computing power available to researchers. TeraGrid’s 11 partner organizations include universities, government labs, and supercomputing centers, and the NSF as well. (TeraGrid resources include more than 750 Tflops of computing power and 30 PB of online and archival storage.)
Clusters, Clouds, Grids, and other versions of our continuously expanding massively parallel supercomputing infrastructure all point to a future more integrated with high-performance computing capabilities than we can possibly even envision. Indeed, the Singularity posed by Vernor Vinge in 1993 as occurring no later than 2030 appears to be more real today than it was fifteen years ago. The Singularity is a point in time where technology accelerates faster than anyone is able to predict what will occur.
As an editor for a science publication, I sometimes feel that I’m jaded by the technologies to which I’m exposed—many of the advanced technologies we wrote about in the past often didn’t see broad acceptance or implementation for several years. Lately, however the advanced technologies we write about are closer to being broadly implemented and more often now are already in use. And obviously, today’s advanced technologies are considerably more sophisticated than those of the past.
