IBM's Watson Blue Gene

Last week IBM unveiled its latest and greatest supercomputer. The
91-teraflop monster, dubbed "Watson Blue Gene," is capable of cranking
out a mind-blowing 91 trillion calculations per second. This ultra-powerful
-- yet remarkably affordable, practical, and accessible -- computing tool is
likely to give IBM a significant competitive advantage over competitors
such as Affymetrix, Intel, and General Electric in the life sciences,
information technology, and material sciences sectors.

The first area where the new computer will prove useful is in modeling
how proteins fold. A number of illnesses, including Alzheimer's disease,
Parkinson's disease, cystic fibrosis, mad-cow disease, and sickle-cell
anemia, are all thought to be related to improperly folding proteins. It's
expected that by better understanding the processes through which these
diseases occur, IBM researchers will find new and better ways to treat
them. Ultimately, the researchers hope to prevent the diseases from ever
occurring.

What it means for IBM is that the company could become not just a
bigger player in the life sciences sector but quite possibly the dominant
one. But the supercomputer's utility is by no means limited to examining
protein folding. Its raw processing power can be applied to any field or
discipline that requires massive crunching power, including financial risk
modeling, climate modeling, seismic exploration, and automotive design,
to name just a few.

As a result of this explosion of new data and information, IBM's consulting
business is likely to reap a lot of new contracts from companies hoping to
profit from the new findings. The computer should also strengthen IBM's
already strong position in the burgeoning field of nanotechnology. By
reducing the time it takes to run simulations from months or weeks down
to mere days or hours, Watson Blue Gene holds the potential to
exponentially increase our understanding of the complex fields of
materials sciences, quantum chemistry, and molecular and fluid dynamics.
Furthermore, because the computer is interactive, researchers will be able
to make adjustments on the fly, and that should speed up the time it takes
to translate scientific discoveries into viable commercial products.

In addition to being super-fast, Watson Blue Gene is also scalable. In fact,
IBM officials are confident that they'll be able to develop a successor
capable of a 1 petaflop performance level by the end of the decade. That's
1 quadrillion calculations per second -- 10 times as powerful as Watson
Blue Gene.