Sometimes my head hurts. It hurts from trying to keep up with the advances in supercomputing. More specifically, it hurts from trying to comprehend how these behemoths will change the business landscape in the near future.
Earlier this year, I noted that IBM (NYSE: IBM ) was working with Rensselaer Polytechnic Institute (RPI) to develop a supercomputer capable of 70 trillion calculations per second. At the time, I mused that it would take a person using a handheld calculator 60 million years to perform what this machine could do in a second. Then, this summer, Big Blue announced that it was working with the National Nuclear Security Administration to install a supercomputer capable of 1 quadrillion (1,000 trillion) calculations per second.
And now, just yesterday, the U.S. government's Defense Advanced Research Projects Agency (DARPA) revealed it had awarded $244 million to IBM and another $250 million to Cray (Nasdaq: CRAY ) to design a new generation of supercomputers. This next generation would be 100 times more powerful than today's supercomputers, but simpler to program and easier to use. (The loser in the competition was Sun Microsystems (Nasdaq: SUNW ) , which was not awarded a contract).
If all goes according to plan, one or both of these companies will have a supercomputer churning out a brain-rattling 10 quadrillion operations per second by 2010. This is 10 petaflops in geek-speak, or if you prefer to use my hand-calculator time analogy, it would take you a mere 8 billion years to do what these computers will be able to do in one second.
I am convinced that IBM is going to be the best blue chip of 2007, but not just because it is a leader in the development and creation of supercomputers. Rather, it is because I believe that IBM's Global Consulting business, working in conjunction with its Center for Business Optimization (whose small staff specializes in applying advanced mathematics to business problems), will be able to first help businesses harness the power of these powerful computers to crunch data and then translate that data into meaningful -- and profitable -- insights.
For example, IBM's Center for Business Optimization recently helped a company that had more than 70,000 SKUs (stock keeping units) with an inventory problem. Normally, finding the best way to manage this number of products would have taken six hours. With a supercomputer, it took 17 seconds.
What this implies in more practical terms is that the company, instead of waiting overnight for results, now gets them immediately. More importantly, because the results are available immediately, the company can now insert different variables into the program to find an even better way to optimize its inventory.
Supercomputers' potential, however, far transcends inventory management. I remain convinced that these machines will also lead to startling scientific breakthroughs in medicine, material science, and nanotechnology. And because IBM is positioned at the confluence of how these findings might be used by business, it will also be able to generate substantial revenue for its consulting businesses.
Interested in other Foolishness about Big Blue?
- The Best Blue Chip for 2007: IBM
- Big Blue Is Watching You
- IBM's Got Serious Game
- IBM Takes Another Small Step
Looking for the next great tech innovator? OurMotley Fool Rule Breakersnewsletter can lead the way. And now you can take the swashbuckling newsletter dedicated to up-and-coming companies in emerging fields for a30-day free spin.