The US Department of Energy’s Pacific Northwest National Laboratory and Hewlett-Packard have announced that PNNL has ordered a $24.5 million HP Linux-based supercomputer.
According to PNNL, the supercomputer should be the world’s most powerful Linux-based supercomputer and one of the top supercomputers in the world when it is fully operational.
Consisting of 1,400 Intel Itanium Family Processors that are code-named McKinley and Madison, the new HP supercomputer will have an expected total peak performance of more than 8.3 teraflops. Calculations that currently take a month to complete will be done in one day on the new system.
Scientists will use the new HP supercomputer to study complex chemical problems that form the basis for new discoveries in areas such as biological systems, subsurface transport, material design, atmospheric chemistry and combustion. The supercomputer also will be vital to better understand systems biology, including structural biology, genomics and proteomics.
Scheduled to be fully operational in early 2003, the massively parallel computer is expected to be more than 30 times faster, have 50 times more disk space and hold 10 times as much memory as PNNL’s current supercomputer, one of the world’s most powerful when installed in 1997.
The new supercomputer is slated to be installed in the Molecular Science Computing Facility within the William R. Wiley Environmental Molecular Sciences Laboratory (EMSL), a DOE scientific user facility at PNNL.
‘As we try to use computational results to replace difficult and expensive experiments, increased computational power is essential,’ said Dave Dixon, EMSL associate director for theory, modelling and simulation. ‘The advanced architecture of the HP supercomputer provides the computational power that will permit us to attain close to peak performance on our key computational chemistry problems.’
The new supercomputer will, for example, enable scientists to study a larger group of hydrocarbons, in particular complex hydrocarbons, which are vital to gaining a complete understanding of the reaction mechanisms involved in hydrocarbon fuel combustion.
Providing data about these reactions currently unavailable through experiments should lead to development of more efficient catalysts and minimised carbon dioxide production.