The saying goes that if you take an infinite amount of monkey’s with an infinite amount of typewriters eventually they will write Shakespeare’s literary masterpiece, Hamlet.
But, what if you took an infinite amount of idle desktop computers and servers and let them talk to each other? It’s called peer-to-peer or distributed computing and according to people like Currid & Company analyst Cheryl Currid, P2P has the capacity to change the Internet as we now know it.
“We are at the beginning of a major change in Internet architecture,” Currid remarked Wednesday during her welcome address at the Summit on Peer-to-Peer Computing in San Francisco. “All of us here are pioneers and we are still charting the course. Together we get to write the rules one more time.”
The three-day summit sponsored by DCI is addressing distribution computing with some of the companies who are early adopters of the technology such as Intel, NEC Systems and Entropia.
“We think this will payoff,” says Intel executive Bob Knighten of the future of P2P. “We’re trying to see what the payoff will be.”
“The funny thing is the only people who can predict the future of P2P are 13-year-old kids,” quipped Currid.
Peer-to-peer has jumped in popularity. A recent Jupiter/Media Metrix survey of Web site hits in January and February said the two sites with the highest traffic in Jan-Feb were the IRS homepage and Napster, a well-known P2P site.
“Napster is possibly the greatest example of beta testing for peer-to-peer computing,” says Currid. “What it proves is that a peer-to-peer network really works.”
Case in point: the day the courts first slapped an injunction against the San Mateo-based Internet song swapping service. Site traffic quadrupled within hours and everything still worked. If Napster’s system were just in one place, Currid says its servers would be taxed to the limit.
Still, the thought of file sharing and breaking though firewalls is a challenge to some and a threat to others.
“File sharing, I think makes IT people nervous,” says Currid. “We saw the same type of reaction with the advent of client/server protocols. I knew IT managers who folded their arms at company meetings saying ‘No, no, hell no We’re not going to do this at my company.'”
And if you still think there’s no market for P2P computing, consider that as much as 75 percent of desktop computers spend their time in idle. Currid speculates if an application could grab their unused cycles and band them together, they could use these computers for processing.
The same would go for hard disks. Just take the unused gigabytes of storage from each unit in a fleet of desktops, and organizations can increase their storage capacity by terabytes.
For example, IBM’s ASCII Big White supercomputer boasts speeds of 12.3 teraflops. Currid says some P2P networks can out pace Big White with ten times the speed.
