TATA’S EKA: The Asia’s Fastest Supercomputer

Digital Super ComputerTata Group’s supercomputer “EKA” literally means ‘one’ in Sanskrit. Eka has been ranked, as the 4th fastest supercomputer in the world, and the fastest supercomputer in Asia, in the Top500 Supercomputer list announced at an International Conference for High-Performance Computing at Reno (California), USA

IBM’s BlueGene/L has grabbed the first position with 478.2 Teraflops, while the second position has gone to the IBM’s newer version of BlueGene/L called BlueGene/P.

The supercomputer uses 14240 Intel’s high-speed (3 giga hertz) ‘quad-core’ Clovertown (Xeon 53xx) processors in nearly 1,800 computing nodes put together on a Hewlett-Packard Cluster Platform 3000 BL460c system. EKA has been developed at a cost of $30 million and has been built in just 20 months.

Eka has benchmark performance of 120 teraflops (trillion floating point calculations or those involving numbers with a floating or decimal point). The supercomputer was developed at Tata Group’s Computational Research Laboratories (CRL) based in Pune.
Incidentally, it’s the first time ever that India has figured in the ‘Top-10 Supercomputer Sites’ list. A total of nine supercomputers developed in India have appeared in the Top 500 list.

Ratan Tata, the group chairman said, “The Tata Group has supported this development actively… I am sure this supercomputer and its successor systems will make a major contribution to India’s ongoing scientific and technological initiatives.”

Ramadorai, Chairman of CRL said, “It is a team effort rather than an individual’s effort. This has put India on the world map and brought a national sense of pride.”

Raju Bhinge, Chief Executive of Tata Strategic Management Group said, “In another six to nine months, we would be able to build applications and a software library, following which we would take the offering to commercial use.”

The supercomputer rankings have been compiled by Hans Meuer of the University of Mannheim, Germany; Erich Strohmaier and Horst Simon of NERSC/Lawrence Berkeley National Laboratory; and Jack Dongarra of the University of Tennessee, Knoxville.