WARF and Intel Settle Patent Suit Over Core 2 Duo

reebmmm writes "The Wisconsin Alumni Research Foundation (WARF) and Intel have settled their patent suit over technology developed by Gurindar Sohi, a computer science professor at the University of Wisconsin — Madison. Professor Sohi developed technology that was ultimately patented by WARF using money he received from Intel. Last month, Judge Barbara Crabb found that the funding agreement was ambiguous, but that e-mails revealed that the money was an unrestricted gift and carried with it no obligation to license or assign any inventions to Intel. Trial was scheduled to begin today. The terms of the settlement were not disclosed."

Re:Hope he never gets funded again

[PalmKiller]

He had a gentleman's agreement to work on the chip it seems from the litigation. Once the university caught on, he tucked his tail and screwed Intel over, so much for the Gentleman part, the a-hole should have owned up to the actual verbal deal with Intel in court, so the suit by the university was null and void, and took his walking papers (and/or lawsuit) from the university like a real man.

Re:Hope he never gets funded again

[Grishnakh]

Yes, but it sucks that people have to have so much legalese in every dealing they have with others, because there's absolutely no trust, respect, or decency left. Just from the description of this story, it seems pretty obvious. Uni gets funding from Intel, develops new tech, patents it, Intel uses it, and gets sued. Please explain how this is correct moral behavior. It isn't. It might be legal, but it isn't right. And this means that tech companies are going to be much less trusting of Universities when thinking about handing out big bags of money to fund research, which is something we desperately need more of in this country to keep on top in technology (since we're losing in everything else, namely manufacturing).

Patent was on memory disambiguation

[DrDitto]

If anyone cares, the patent deals with memory disambiguation. The basic jist is that it is hard to execute *memory* instructions out-of-order when previous the address computation of previous instructions has not completed (otherwise what would happen if the processor completes a load instruction, out-of-order, for a prior store instruction that did not yet complete due to a dependence on address computation?). Sohi's patent figured out a way to predict this and to allow the Core2 to get much better out-of-order execution.

Sohi is *highly* respected in the field of computer architecture. In fact Wisconsin is considered one of the best computer architecture schools in the world.