Blue Gene supercomputer to be 'smashingly' equipped for huge modelling task of protein folding

Proteins control all cellular processes in the human body. Comprising strings of amino acids which are joined like links of a chain, a protein folds into a highly complex, three-dimensional shape that determines its function. Any change in shape dramatically alters the function of a protein, and even the slightest change in the folding process can turn a desirable protein into a disease.

Better understanding of how proteins fold will give scientists and doctors better insight into diseases and ways to combat them. Pharmaceutical companies could design high-tech prescription drugs customised to the specific needs of individual people. Physicians could respond more rapidly to changes in bacteria and viruses which cause them to become drug-resistant.

About 50 scientists from IBM Research's Deep Computing Institute and Computational Biology Group are working on Blue Gene and the protein folding grand challenge, as Dr. Pulleybank explained. The Blue Gene supercomputer will cost some 100 million US dollars and will be 500 times more powerful than the world's fastest computers today. The system will be capable of more than one quadrillion operations per second, representing one petaflop. This level of performance will make Blue Gene a 1000 times more powerful than the Deep Blue machine which beat world chess champion Garry Kasparov in 1997, and about 2 million times more powerful than today's top desktop PCs.

Dr. Pulleybank showed how Blue Gene's massive computing power will initially be used to model the folding of human proteins. The experimental new architecture forms the key to petaflop performance. Dr. Pulleybank anticipated that this radically new approach to computer design and architecture will enable Blue Gene to achieve petaflop-scale performance in about five years. This is one third of the close to 15 years it would normally take following Moore's Law. The two fastest computers in the world at present are part of the ASCI programme run by the United States Department of Energy. They both were recently tested at about 2 teraflops or two trillion operations per second each.

The Blue Gene project team expects a tremendous gain in performance will be made possible by the first major revolution in how computers are built since the mid-1980s. This new approach to computer architecture is called SMASH, which stands for Simple, Many and Self-Healing. The innovative SMASH architecture differs from existing approaches in three ways. First of all, it dramatically simplifies the number of instructions carried out by each processor, allowing them to work faster and with significantly lower power and chip surface requirements. The traditional approach consisted in adding complex features to gain performance.

Second, it will facilitate a massively parallel system capable of more than 8 million simultaneous threads of computation, which is tremendous compared to the maximum of 5000 threads today. Third, it will make the computer self-stabilising and self-healing, meaning that the Blue Gene system will automatically be able to overcome failures of individual processors and computing threads. Blue Gene will consist of more than one million processors, each capable of one billion operations per second, which amounts to 1 gigaflop.

Thirty-two of these ultra-fast processors will be placed on a single chip, thus hosting 32 gigaflops. A compact two-foot by two-foot board containing 64 of these chips will be capable of 2 teraflops, making it as powerful as the 8000-square foot ASCI computers. Eight of these boards will be placed in 6-foot-high racks, representing 16 teraflops, and the final machine with a volume of less than 2000 square feet will consist of 64 racks linked together to achieve the one petaflop performance.