Drug maker Vertex Pharmaceuticals Inc. has commissioned the construction of a 112-processor, 110-gigaflop Linux supercomputer cluster that will be used to accelerate drug research and development.
In an announcement, Cambridge, Massachusetts-based Vertex said the hardware is being provided by VA Linux Systems Inc. in Fremont, California, and the project management, software deployment and configuration are being provided by Boston-based Blackstone Computing.
The cluster will initially use 56 computers, each running a pair of Intel Corp. 933-MHz Pentium III processors, and will be scaleable to up to 450 machines running as many as 900 processors. The cluster will perform parallel computations using proprietary software to conduct Vertex research into structural biology, combinatorial chemistry, medicinal chemistry, bioinformatics and pharmacology groups.
The company said the cluster represents a key component of the technology infrastructure that will accelerate Vertex's gene family-based research.
Andy Fant, a Vertex senior systems engineer, said the new machine will replace a 4-year-old 45- to 50-megaflop Silicon Graphics Inc. computer system that is approaching the end of its lease. The new system is still being set up and tested, Fant said, and won't go fully online for several weeks.
The Linux system has an easy-to-use interface for Vertex scientists, he said. "It's very important for our scientists who don't want to become computer jockeys."
The exact price of the cluster isn't being disclosed, but it's more than US$500,000, according to a spokesman.
Similar Linux-based supercomputers have already been announced this year.
Two weeks ago, the National Science Foundation announced a $53 million project to connect a series of remotely located powerful computers into a high-speed Linux supercomputer grid that could open vast new opportunities for scientific and medical breakthroughs.
In March, the U.S. Department of Defense announced it would build a 512-processor Linux cluster at a computing facility in Hawaii. That cluster will be able to process 478 billion calculations per second for applications such as tracking and fighting wildfires across the country.