Intel in cooperation with IBM has created a computer chip that delivers a teraflop of performance. That means the chip is able to perform a trillion floating point mathematical operations per second. At the moment, it takes a supercomputer or a computing cluster to achieve that kind of performance. The performance was achieved by figured out a way to put 80 core calculating engines on a single chip. The chip consumes much less power than other similar available methods (62 watts at 3.16 gHz vs 500 kilowatts for 10,000 pentium chips). At the same time the chip contains about 1/3 less transistors than Intel's previous chip; or about 100 million transistors.