IBM will now use Linux operating systems in its Blue Gene line of
powerful supercomputers.
The company said the open source Linux operating system was
selected because it offers flexibility and has a large base of
developers and contributors to help solve challenges as they arise.
"We had two choices of operating systems for the Blue Gene family,
either use a special purpose system or Linux," said Bill
Pulleyblank, director of Exploratory Server Systems at IBM
Research. "We saw considerable advantage in using an operating
system supported by [the] open source community so that we can get
their input and feedback."
Takako Yamakura, a spokesman for IBM Research, said the early focus
in the Blue Gene project was on hardware. But as Blue Gene has
evolved, the operating system has received more attention from
engineers - leading to the official designation by IBM of Linux as
the operating system of choice.
The most recent addition to the Blue Gene family is the Blue Gene/L
supercomputer being built by IBM and the Lawrence Livermore
National Laboratory for the US Department of Energy's National
Nuclear Security Administration. The machine will be at least 15
times faster and 15 times more power-efficient than today's fastest
supercomputers, but will take up only a fraction of the space.
The original Blue Gene supercomputer is expected to be complete by
2004, while Blue Gene/L is expected to be ready the following
year.
Blue Gene/L is expected to operate at about 200 trillion
floating-point operations per second - larger than the total
computing power of the top 500 supercomputers in the world today,
according to IBM. Blue Gene/L will also include IBM systems that
allow complex computers to do self-repairing, self-managing and
self-configuring, making them easier to manage and set up.
For IBM, the latest Blue Gene/L project is designed to demonstrate
commercial uses for the powerful machines so the company can market
that end of its research for additional sales.