Supercomputing from your PC
June 27, 2005
By Canadian Consulting Engineer
Your personal computer sitting idly at home can now become part of a vast dispersed supercomputer network solving c...
Your personal computer sitting idly at home can now become part of a vast dispersed supercomputer network solving complex scientific problems. Dispersed virtual networks are being used for work such as uncovering Einstein’s gravitational ripples or finding drug compounds to fight AIDS. According to an article by Charles Q. Choi in Scientific American magazine published May 16, there are at least 60 so-called “@home” projects on the Internet where PC-users can donate their processing power to scientific projects like the recently launched Einstein@home.
The typical modern PC performs about one-billion floating-point operations per second but aside from graphical programs, generally only uses a fraction of this power. The @home projects use this spare PC power in a process called distributed computing. While IBM’s largest supercomputer BlueGene/L has a 70-trillion flop capacity, SETI@home running off 500,000 PCs has 100-trillion flops.
According to the Scientific American article, the surge in distributed computing comes from the development of platforms such as the Berkeley Open Infrastructure for Network Computing that allow for a computer to host multiple projects.
The first public distributed computing project was the Great Internet Mersenne Prime Search, launched in 1996 to look for large prime numbers. Since then have been projects such as the FightAIDS@home and the Monkey Shakespeare Simulator.