Sign in with
Sign up | Sign in
Your question

Opti Files Suit Against AMD?

Last response: in CPUs
Share
November 19, 2006 3:17:19 AM

http://www.internetnews.com/bus-news/article.php/3644786
Quote:
Opti announced it has filed a patent infringement lawsuit on Friday against AMD over three patents under the umbrella name of "Predictive Snooping of Cache Memory for Master-Initiated Accesses."

In a release, Opti said AMD (Quote) infringes the patents by making, selling, and offering for sale CPUs and core logic products based on and incorporating Predictive Snooping technology and inducing and contributing to the infringement of the patents by others.

The suit was filed the U.S. District Court for the Eastern District of Texas.

An AMD spokesperson said the company had not seen the suit and had no comment at this time.

But Bernard Marren, president & CEO of Opti, said the company first approached AMD four years ago.

"Back then, they were only using [the technology] in their chipsets, not in big volume," Marren told internetnews.com. "This is a more serious case because it's across the board," he said, referring to AMD's Opteron processors.

Mountain View, Calif.–based Opti is mainly focused on licensing patents it attained during its years as a chipset designer, which it stopped doing about five years ago.

According to Merran, Opti settled a similar case, involving the same patents, with nVidia (Quote) in August for $11 million and a royalty fee of $750,000 per quarter starting next year. He also said Opti struck a cross-licensing deal with Intel (Quote) years ago that netted the firm $13.5 million and protects Intel from being sued in this matter.


But does it have any legs? This conjures up thoughts of Transmeta v Intel.

More about : opti files suit amd

a b à CPUs
a b À AMD
November 19, 2006 11:49:53 AM

It will be a kick in the goolies for AMD if it does. Then again they may just liquidate recently acquired assets to settle out of court. :twisted:
November 19, 2006 9:21:20 PM

Quote:
http://www.internetnews.com/bus-news/article.php/3644786
Opti announced it has filed a patent infringement lawsuit on Friday against AMD over three patents under the umbrella name of "Predictive Snooping of Cache Memory for Master-Initiated Accesses."

In a release, Opti said AMD (Quote) infringes the patents by making, selling, and offering for sale CPUs and core logic products based on and incorporating Predictive Snooping technology and inducing and contributing to the infringement of the patents by others.

The suit was filed the U.S. District Court for the Eastern District of Texas.

An AMD spokesperson said the company had not seen the suit and had no comment at this time.

But Bernard Marren, president & CEO of Opti, said the company first approached AMD four years ago.

"Back then, they were only using [the technology] in their chipsets, not in big volume," Marren told internetnews.com. "This is a more serious case because it's across the board," he said, referring to AMD's Opteron processors.

Mountain View, Calif.–based Opti is mainly focused on licensing patents it attained during its years as a chipset designer, which it stopped doing about five years ago.

According to Merran, Opti settled a similar case, involving the same patents, with nVidia (Quote) in August for $11 million and a royalty fee of $750,000 per quarter starting next year. He also said Opti struck a cross-licensing deal with Intel (Quote) years ago that netted the firm $13.5 million and protects Intel from being sued in this matter.


But does it have any legs? This conjures up thoughts of Transmeta v Intel.

In a somewhat naïve way, here's my verdict:

As you've mentioned, nVidia also incurred in OPTi's patents violations (http://news.zdnet.co.uk/itmanagement/0,1000000308,39170845,00.htm):
Quote:
In 2002, OPTi sold its semiconductor business to OPTi Technologies. At that time, Marren said that OPTi would turn its attention to protecting and enforcing its intellectual property rights.

If one's to take into account the way Marren's argument is put forward:
Quote:
But Bernard Marren, president & CEO of Opti, said the company first approached AMD four years ago.
"Back then, they were only using [the technology] in their chipsets, not in big volume," Marren told internetnews.com. "This is a more serious case because it's across the board," he said, referring to AMD's Opteron processors.

And finally, what is OPTi, nowadays:
http://www.opti-inc.com/html/products.html

I'll dare to speculate that this litigation will be easily settled, for both parts:

First off, nVidia isn't struggling to survive, due to the lawsuit following the infringements;
Secondly, Marren's argument (as quoted) concerning AMD's patents infringements is weak, at best; Marren seems to try to reinforce Opti's position, not because AMD prevaricated before (chipsets) but rather, because it's using the technology in a highly profitable, successful technological vector: CPUs.
Finally, OPTi Technologies, Inc. is no-longer a contender, in the chipset arena; its IP assets' worth are easier to negociate now, as it's of nobody's interest that "Predictive Snooping of Cache Memory for Master-Initiated Accesses" becomes doomed to oblivion, either by legal or illegal lack of use.

Anyway, my naïve point-of-view. :wink:


Cheers!
Related resources
November 19, 2006 9:30:21 PM

I'd thought as much. It seemed to me to be an oppertunistic case in all senses of the term.
By the way, "Predictive Snooping of Cache Memory for Master-Initiated Accesses" is an incredibly long name for what technology?
November 19, 2006 9:32:14 PM

Quote:
http://www.internetnews.com/bus-news/article.php/3644786
Opti announced it has filed a patent infringement lawsuit on Friday against AMD over three patents under the umbrella name of "Predictive Snooping of Cache Memory for Master-Initiated Accesses."

In a release, Opti said AMD (Quote) infringes the patents by making, selling, and offering for sale CPUs and core logic products based on and incorporating Predictive Snooping technology and inducing and contributing to the infringement of the patents by others.

The suit was filed the U.S. District Court for the Eastern District of Texas.

An AMD spokesperson said the company had not seen the suit and had no comment at this time.

But Bernard Marren, president & CEO of Opti, said the company first approached AMD four years ago.

"Back then, they were only using [the technology] in their chipsets, not in big volume," Marren told internetnews.com. "This is a more serious case because it's across the board," he said, referring to AMD's Opteron processors.

Mountain View, Calif.–based Opti is mainly focused on licensing patents it attained during its years as a chipset designer, which it stopped doing about five years ago.

According to Merran, Opti settled a similar case, involving the same patents, with nVidia (Quote) in August for $11 million and a royalty fee of $750,000 per quarter starting next year. He also said Opti struck a cross-licensing deal with Intel (Quote) years ago that netted the firm $13.5 million and protects Intel from being sued in this matter.


But does it have any legs? This conjures up thoughts of Transmeta v Intel.

They settled with Intel and Nvidia they will settle with AMD, might be a bit more pricey now with regards to the 4 year statement, but well all have to wait and see.
November 19, 2006 10:49:32 PM

Quote:
I'd thought as much. It seemed to me to be an oppertunistic case in all senses of the term.
By the way, "Predictive Snooping of Cache Memory for Master-Initiated Accesses" is an incredibly long name for what technology?


Good question; I was already trying to search for some info on it; much better than I, by the authors, themselves:

Quote:
Abstract

When a PCI-bus controller receives a request from a PCI-bus master to transfer data with an address in secondary memory, the controller performs an initial inquire cycle and withholds TRDY# to the PCI-bus master until any write-back cycle completes. The controller then allows the burst access to take place between secondary memory and the PCI-bus master, and simultaneously and predictively, performs an inquire cycle of the L1 cache for the next cache line. In this manner, if the PCI burst continues past the cache line boundary, the new inquire cycle will already have taken place, or will already be in progress, thereby allowing the burst to proceed with, at most, a short delay. Predictive snoop cycles are not performed if the first transfer of a PCI-bus master access would be the last transfer before a cache line boundary is reached.


http://www.patentstorm.us/patents/5813036-fulltext.html

I'll need sometime to digest it, though.


Cheers!
November 19, 2006 10:56:26 PM

So basically this technology is integral to AMD's method of building CPU's if I am at all correct? The patent language seems quite convoluted to say the least. Or is it a secondary method for accessing the PCI bus? :?
November 20, 2006 12:27:58 AM

Quote:
So basically this technology is integral to AMD's method of building CPU's if I am at all correct? The patent language seems quite convoluted to say the least. Or is it a secondary method for accessing the PCI bus? :?


Well, it seems part of cache snooping & data pre-fetching, rather an adaptation of the technology than a copy/paste; it seems to have something in common with memory disambiguation as well, but I confess my ignorance on this subject. I'd need more time to go through it.
Perhaps Ltcommander_data or Jumping Jack could come up with a more straightforward synthesis...

I'll get back to you, if I can digest it properly.


Cheers!
November 20, 2006 1:15:00 AM

I'd be much obliged to you.
!