WARF raiders slap Intel with Core patent suit
Wisconsin silicon heads
The patent defense arm of the University of Wisconsin hit Intel with a lawsuit this week, claiming abuse of intellectual property related to parallel processing.
The Wisconsin Alumni Research Foundation (WARF) has accused Intel of muscling in on US patent 5,781,752 which was granted with the title "Table Based Data Speculation Circuit for Parallel Processing Computer." The patent appears to cover a method of speculation where some advanced instructions are executed before others "on which they may be data dependent." By leaping ahead with some instructions, the processor can improve overall performance and handle multi-threaded software well.
The lawsuit claims Intel received a briefing on the technology from a Wisconsin researcher while the patent application was pending. In addition, the Wisconsin inventors allegedly continued to discuss licensing options with Intel as their patent moved through the approval process.
"Intel refused attempts to license the technology . . .," the lawsuit states. "Rather, Intel incorporated the work . . . into its planning of future products."
According to WARF, the technology in question has been found in Intel's Core micro architecture used as the basis for most of Intel's current chips.
WARF is seeking a jury trial and the standard compensation for damages.
"We’ve been served with the suit and are currently evaluating the issue," said an Intel spokesman. "I can confirm that we were in discussions with WARF on this issue for more than a year. We would expect to file our response in about 30 days."
In the past, WARF has enjoyed some success pursuing similar cases. It reached an out of court settlement with IBM in 2005 over technology related to copper-based computer chips. ®
COMMENTS
@ Scott
Must state-owned corporations share their profits directly with the public? Of course not. Universities should be considered no differently, and there is nothing preventing them from owning a corporate asset. Anyone can bequeath an asset -- including a corporation -- to a university. WARF may not fit this description, but the situation is no always black and white. And as John Stirling above said, sometimes increasing IncomeSource2 can reduce the burden on IncomeSource1. If I, the taxpayer, am IncomeSource1, then I heartily welcome the exploitation of IncomeSource2!
patents & the public institutions
patents are not a natural right. They are a limited right to exclusivity granted by law to allow the patent holder (usually the investor/discoverer) to benefit from their innovation. The reason they exist as a legal reality is to incentivise the sort of research that leads to new discoveries.
You could argue that public bodies do not need to patent discoveries, as they are publicly funded, however the problem with this is that then private corporations simply perform a land grab, patent the stuff themselves, and sue the hell out of anyone who tries to use the method in question. So for this reason public institutions do need to patent discoveries, even if only as a defensive measure.
When you also consider that quite a number of the big successful behemoths that bestride silicon valley started up as university funded or created companies, or were based on tech developed in collaboration with universities (and so with an element of public funding) then it gets even more complicated.
In all probability the current length of patents is probably a little too long, and it is certainly arguable that in the fast moving world of tech especially having them expire sooner would serve no commercial damage to innovative companies, and provide considerable public interest, as it would at least mitigate the patent portfolio licensing companies, if not destroy them.
But to argue that bodies that are publicly funded deserve no legislative protection shows more of a philosophical position than a well reasoned argument (and none the worse for that). If public bodies raise revenue from other sources than taxation then the tax burden can decrease (or more services can be provided). If this is through some form of commercial activity (for example research leading to the granting of patents), then provided competition is not unfair, it is simply another form of taxation, targeted at those who benefit.
Of course the same argument applies to copyright - with bells on.
Prior Art M'lud.
Ever since the mid eighties ARM coprocessors have opeated on the principal of "indempotency", in that a pipeline could execute instructions and hold results, but couldn't commit them back to the register bank until the controlling pipeline approved the instruction.
Obviously there are slight differences in that this patent relates to core pipelines rather than coprocessor pipelines, but they are both examples of interlocked execution units, and as both core and coprocessor register banks can be considered to be subsets of the "CPU state" these differences are trivial to overcome.
The case I am describing has the distinct advantage that hardware incorportaing this principle has existed since at least 1988!
So yes, this is another dumbass patent dispute... Surely the relevent offices have enough work to do debunking bogus applications without smartarses filing patents "just to show how broken the system is"?
@ All
The University of Wisconsin is funded in part by federal taxes. While it's probably true that this is not the university's major source of funding, it still probably represents a sizable piece of its funding. These taxes are paid by every citizen and business in this country. Therefore, every business and citizen in the U.S. who so wishes should be able to make use of the results of their research (after all, we all helped pay for it -- whether or not we wanted to).
Whether everyone in the U.S. is capable of making use of this information is immaterial. Anyone who can make use of the information should be able to. Besides, patents are good for a really fsking long time, and there may be one day in the future (however unlikely it may seem now) when fabrication technology advances to the point where people will be capable of creating their own microprocessors "in their garages" so to speak.
@amanfrommars, RE: Private Institutions
You're not making any sense, man. I'd love to be able to debate you on this, but I haven't a clue what you're trying to say.
Depends on the patent details.
Feck. Enter key...
From the complaint, there's nothing that actually tells you how to USE this technology. Does it tell you the algorithms you need to implement to choose? What weighting to give to results to reduce the impact of getting it wrong?
If it has neither of these, the patent is useless. Rather like a patent saying "trapping mice" and then suing every mousetrap in existence. You can't build my mousetrap with that patent, because I don't have one.
If there is enough detail, then it's a proper patent.
However, this is almost entirely algorithms. And aren't algorithms supposed to be unpatentable???
