August 22nd, 2008
Build your own cloud with open source
For many corporate researchers, clouds and distributed computing have a big downside. (This cluster, at the University of Texas, is looking for oil reservoirs, naturally.)
They’re somewhere else. They’re someone else’s hardware.
If you’ve got a big job, with enormous financial implications, you want to run it on your own systems. But how?
With open source tools from REvolution Computing. With backing from Intel Capital and a board of technical advisors that has paid its Larry Augustin tax, Richard Schultz is looking for corporate clusters whose big database sorts might cure cancer.
Actually, his first niche is in genetics, where scientists at firms like Pfizer or Novartis want to sort massive databases to find compounds with very specific properties.
The open source tools are R, a language designed for such sorts, and NetWorkSpaces, the “secret sauce” which lets R run in a cluster or cloud.
Actually, the sauce is not so secret.
Both tools are open source. “We are dedicated to making sure the secret sauce remains state of the art,” said Schultz. “A small number of developers is quite impractical these days.”
Open source, in other words, is the secret sauce that makes the REvolution happen.
Dana Blankenhorn has been a business journalist for 30 years, a tech freelancer since 1983. See his full profile and disclosure of his industry affiliations.
| 08/22/08
