IBM building first ‘self-aware’ supercomputer

Supercomputing leader IBM Corp. on Friday announced that it has begun assembling a colossal supercomputer called Blue Sky for The National Center for Atmospheric Research (NCAR), in Boulder, Colorado.

Capable of predicting atmospheric climate changes, heating oil prices, and global warming, Blue Sky will be equipped with IBM’s eLiza technology by the end of next year. The goal of IBM’s eLiza program is to give a computer the ability to repair itself, and keep itself running without human intervention.

The first stage of Blue Sky’s assembly at NCAR, code-named “Black Forest,” will line up over 300 IBM SP Supercomputers to deliver computing power equal to 2 trillion calculations per second, according to Peter Ungaro, the vice-president of high performance computing at IBM.

The second phase, scheduled to begin in the Fall of 2002, will attach a mammoth array of IBM’s p690 Unix servers to the SP systems, yielding a peak performance of 7 trillion calculations per second, said Ungaro.

The p690 servers will be equipped with IBM’s recently introduced eLiza “self-healing” technology, making the NCAR supercomputer the first eLiza supercomputer built outside of IBM’s labs.

“These p690 systems will have eLiza in them, and will be capable of doing diagnostics on themselves, keeping themselves reliable,” said Ungaro. “From our view, we try and build computers that don’t go down and putting self-healing technology in them helps us achieve that.”

While Blue Sky is not the biggest supercomputer in the world, its eLiza technology will make it one of the most self-sufficient.

Supercomputing experts have long predicted that supercomputing systems will eventually become so fast and so complex, that trouble-shooting them will be only possible through the use of another supercomputer. This scenario is best described in a paper by Bill Joy, the chief scientist of Sun Microsystems, titled “Why the future does not need us.”

When asked if IBM supercomputers like Blue Sky were reaching performance levels that exceeded an IT staff’s ability to accurately de-bug them, Ungaro said “that’s a difficult question to answer, but that’s what eLiza is for.”

IBM’s biggest system is a 12.3 teraflop, non-eLiza supercomputer based at the Lawrence Livermore Laboratory, in Livermore, California, said Ungaro.

How big IBM can go with a supercomputer is, surprisingly enough, just a matter of economics, said Ungaro.

“Right now we believe we can build a bigger supercomputer than anyone has the budget to buy,” he said. “We don’t see a ceiling.”

Ungaro does predict however that within five years, IBM should be assembling a supercomputer capable of calculating in petaflops. A supercomputer of this size would be faster than all 500 of the world’s fastest supercomputers, multiplied by ten, he said.