Whether or not you follow news regarding Apple Computer, you have no doubt heard of the Virginia Tech 1,100 node G5 supercomputer. I’m one lucky dog. I’m writing a book about the project, and the folks over at Virginia Tech have treated me like part of the team. I’ve been privileged to watch the construction of the project and I can honestly tell you it left me speechless.
On my first visit, the machines were all unpacked and in the racks, but shrouded within sheets of plastic, to keep them safe from dust caused by the construction. It gave the room an erie look. All around me, construction crews were working to install the cooling and power systems. Cables littered the floor and teams were busy running them to each machine.
On my second visit, the plastic was gone, the cables were complete, and we started to fire up all 1,100 machines. I even had the thrill of turning on six of the machines myself. We joked about kicking off the power and turning it back on just to hear the din of the system chime from 1,100 Macs.
This project is important on many levels. Obviously this is an important step forward for Apple’s entry into the Enterprise market. It is hard to dispute the performance of Macs when you have a G5 cluster outperforming many of todays fastest supercomputers and doing it with fewer machines. On top of that, to do it for less money than any other platform available today, goes against everything the Windows/PC world wants you to believe about the price/performance factor of Apple hardware. The decision for the folks over at Virginia Tech was simple, get the best performance for the best price. They weren’t the only people shocked when that turned out to be Apple…Apple was a little surprised as well…I mean hey…when people built supercomputers, nobody called Apple…until now.
So, will this change anything for Apple? Will it gain new respect for the company at 1 Infinite Loop? More importantly, will it translate to sales? No one can say for sure. There is no doubt a prejudice against Apple amongst many IT departments. In the past, some of that prejudice was deserved. However, Apple is on a roll these days. Design awards, speed boosts, awards for OS X, a new respect from Unix geeks…and companies who never considered porting their software to the Mac platform, are taking a second look.
This story dates back to last year. It almost never happened. But it is far from over. Next year, all the little known details will be revealed in a new book. By that time we’ll know what the project means for supercomputing and for Apple.
If you want to learn more about the project, there will be a keynote speech from one of the project leaders, Dr. Srinidhi Varadarajan, at the O’Reilly OS X conference.
Also, on my last visit, the BBC was there filming a news piece on the story. You can find that here.