Beowulf is an approach to building a supercomputer as a cluster of commodity off-the-shelf personal computers, interconnected with a local area network technology like Ethernet, and running programs written for parallel processing. The Beowulf idea is said to enable the average university computer science department or small research company to build its own small supercomputer that can operate in the gigaflop (billions of operations per second) range. In addition to possible cost savings, building your own supercomputer is said to be a learning investment and make you less dependent in the future on particular hardware and software vendors. As off-the-shelf technology evolves, a Beowulf can be upgraded to take advantage of it.
By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
The original Beowulf cluster was developed in 1994 at the Center of Excellence in Space Data and Information Sciences (CESDIS), a contractor to the US National Aeronautics and Space Administration (NASA) at the Goddard Space Flight Center in Greenbelt, Maryland. Thomas Sterling and Don Becker built a cluster computer that consisted of 16 Intel DX4 processors connected by channel-bonded 10 Mbps Ethernet. Their success led to the Beowulf Project, which fosters the development of similar commodity off-the-shelf (COTS) clusters. A number have been developed in universities and research groups, ranging from the original 16-processor Beowulf to Avalon, a cluster of 140 Alpha processors built by the Los Alamos National Laboratory. A more typical smaller cluster might have 16 200-MHz (or faster) Intel P6 processors connected by Fast Ethernet and a Fast Ethernet switch.
As a way to lower cost and increase vendor independency, Beowulf developers often choose the Linux operating system and use standard message passing protocols between the computers within the cluster. A Beowulf cluster is placed in the taxonomy of parallel computing as somewhere below a massively parallel processor (MPP) and a network of workstations (NOW) that is clustered for the purpose of load balancing.