To: StanX Long who wrote (48317 ) 6/21/2001 4:36:43 PM From: Proud_Infidel Respond to of 70976 IBM supercomputer to soup up cars By The Associated Press Special to CNET News.com June 21, 2001, 12:30 p.m. PT NEW YORK--When it comes to innovations in speed and efficiency, cars are no match for computers. In the past few decades, autos have shown little improvement in performance per gallon of fuel consumed. In the same period, computing performance has leaped more than a hundredfold without similar increases in use of electricity. Now, computers are being tapped to help the languishing cars. Thursday, at a supercomputing conference in Germany, IBM will announce that it has built the world's second-most powerful supercomputer, which will be dedicated to researching fuel efficiency in automobiles and other energy-related problems. The machine, theoretically capable of as many as 3.8 trillion calculations per second, will be installed at the National Energy Research Scientific Computing Center (NERSC) at the Energy Department's Lawrence Berkeley National Laboratory in Oakland, Calif., IBM announced. The computer will be used to test computer models of internal combustion engines in hopes of finding a model that burns less fuel and emits fewer pollutants, said NERSC spokesman John Hules. The NERSC machine will also be used in global climate modeling and research into fusion energy, proteins, the environment and biology, said Hules. IBM builds more of the world's fastest supercomputers than any other company: 201 of the 500 fastest supercomputers are the product of Big Blue, according to the Top 500 List, maintained by researchers at the University of Tennessee and Germany's University of Mannheim. The No. 2 vendor, Sun Microsystems, counts 81 machines on the list. In terms of computing power, the NERSC model trails only one existing machine: the Accelerated Strategic Computing Initiative White IBM supercomputer installed at the Energy Department's Lawrence Livermore National Laboratory in California. That machine is engaged in classified modeling of nuclear weapons explosions, said Hules.