DailyTech - Rambus Wins $306.9M in Patent Infringement Case

Submit News

Hardware Rambus Wins $306.9M in Patent Infringement Case
Samad Khan (Blog) - April 25, 2006 10:32 AM

33 comment(s) - last by Tanclearas.. on Apr 27 at 8:36 AM

Rambus has won its case against Hynix, with the court ordering Hynix to pay $306.9M

Rambus ’ allegation that Hynix and other large memory manufacturers infringed on its patents in producing DDR DRAM, SDRAM and DDR2 DRAM memory techniques. The vast majority of the PCs and servers made during the last several years have used these types of memory.

The jury on this case determined that Hynix had infringed on 10 of Rambus’ patents. The verdict was announced on Monday in U.S. district court in San Jose, Calif. The award covers compensation for Hynix sales in the United States between June 2000 and the end of 2005, according to a statement by Rambus.

Rambus will seek an injunction barring Hynix from selling DDR, DDR2 and SDRAM this summer. Defendants often agree to pay additional amounts to the plaintiffs through a license to avoid facing the injunction. Rambus also has suits pending against Samsung Electronics, Nanya Technology, Micron Technology and this verdict can prompt Rambus to file even more lawsuits.

Comments

Threshold

Username
Password
remember me

This article is over a month old, voting and posting comments is disabled

It's amazing...

By Ice9 on 4/25/2006 2:35:10 PM , Rating: 5

It's amazing how little people know about this suit.

First, Rambus created the specification for RDRAM back in the 1980's - long before JEDEC ever did any kind of ratification or voting on SDRAM or DDR.

When Rambus innovated these technologies, they filed for patents since it was their idea... That's what rambus does, they are a technology company.

They then showed several large memory manufacturers their idea in the interests of having them make it. Rambus doesn't want to be in the manufacturing business. When they showed these memory manufacturers their ideas, they did so under NDA.

These memory manufacturers then invited Rambus to join JEDEC. Rambus did so because they wanted their new memory to be the standard for all memory technologies used in PC's. In the history of JEDEC, only ONE company was ever barred from presenting their technology for consideration of standardization. That company? Rambus.

Rambus was under no obligation to reveal their patents at JEDEC. Everyone knew about Rambus technology because of the NDA's. But instead, what the steaming pile of JEDEC morons did was cherrypick all of Rambus' patents during the SDRAM standardization process. Rambus saw what they were doing and withdrew from JEDEC. When they left, they clearly stated they were persuing royalties on their technology.

So basically, to those of you thinking you haven't been using Rambus technology in years, think again. If you have ever used SDRAM, DDR, DDR2 or GDDR3, then you are using innovations created by Rambus back in the late 80's. This lawsuit was a trial by jury, and the jury found Hynix guilty - just like the rest of those JEDEC morons will be found guilty.

What the JEDEC companies did was outright theft of Rambus's intellectual property, and there isn't a jury in the world who isn't going to find otherwise.

RE: It's amazing...

By patentman on 4/25/2006 3:32:27 PM , Rating: 2

I agree with you, but I have one word that makes the district court decision relatively meaningless, at least for the short term. That word is "appeal." There is a strong likelihood this case will be appealed to the Federal Circuit Court of Appeals, who are notorious for seeing things in a different light then a district court.

Also, I haven;t read the case, so I dontt know, but are you sure this was tried before a jury? Most patent cases are not tried before a jury because it is hard enough trying to get one non-technically oriented judge to understand the technology involved, much less 12 random people.

Good post though :)

Parent

RE: It's amazing...

By brownba on 4/25/2006 3:36:31 PM , Rating: 2

I found this at bloomberg.com

quote:
Furniss said the jury had found Hynix infringed all eight claims, or elements, of the six patents at issue, and that all the patents were valid. U.S. District Judge Ronald Whyte had previously ruled Hynix infringed two of the claims, bringing the total number of infringed claims to 10.

Parent

RE: It's amazing...

By Lonyo on 4/25/2006 3:32:52 PM , Rating: 3

Right, they created the spec in the 80's.

http://www.rambus.com/about/history.aspx
March 1990 Rambus is incorporated

Even though they have no historical milestones on their own website before 1990....?

They made technology before they existed, they ARE special.

Parent

RE: It's amazing...

By Griswold on 4/25/2006 5:08:11 PM , Rating: 2

Busted!

Parent

RE: It's amazing...

By mushi799 on 4/25/2006 5:29:11 PM , Rating: 3

incorporated in 1990 is different than existing. It took my brother 2-3 years before his business became official, it doesn't mean it was nonexistent

And people forget, that JEDEC and some of its fellow participants was caught for fixing DRam prices. Not only that, but they were caught for forcing Rambus out of the market. Yet Rambus is still the "bad guy".

It's about time Rambus sues these bastards for bad business practice.

Parent

RE: It's amazing...

By atridi on 4/26/2006 4:53:54 PM , Rating: 2

Mark Horowitz and Mike Farmwald developed the technology behing their original patents and RDRAM in a garage not far from Stanford.

You think when Steve Jobs and Steve Wozniak had already formed their corporation when they were working on their first kit computer in the 70's. Hell, if Woz hadn't been addicted to fruit, they'd probably still be trying to come up with a name for the company (and the computer)

Paperwork takes time and costs money. Wake up!

Parent

RE: It's amazing...

By Tuan Nguyen on 4/26/2006 6:42:07 PM , Rating: 2

Actually it was Steve that was the fruitarian. :)

Tuan

Parent

By hoppa on 4/25/2006 12:10:09 PM , Rating: 2

Companies like this make me cry. They do nothing but impede progress across the world for their greedy little CEOs. Why can't they spend their money (errrr, their competitors money) on developing new technology instead of just paying off lawyers to beat up their foes? I don't care if they have a patent on some single transistor that randomly found its way into much new technology, and that they should be defending that. I'm sure all the execs are still rolling in cash and the company would be doing fine without all the lawsuits.

RE: .

By Phynaz on 4/25/2006 1:29:48 PM , Rating: 2

Dude, they DID spend their money on new technology. Now they want to get paid for it.

Parent

RE: .

By Tanclearas on 4/26/2006 10:28:09 AM , Rating: 2

Actually, they spent their money sending moles into JEDEC meetings to snag the best ideas, creating patents that were extensions/modifications to those stolen ideas, and paying lawyers to sue the very people they stole the ideas from.

Parent

RE: .

By atridi on 4/26/2006 4:21:14 PM , Rating: 2

No, actually, they spent their money developing revolutionary memory technology that's in every computer-based product made today. That was in 1989-90. The memory makers then spent the next 5 years trying to use the auspices of JEDEC to steal that technology. NOTHING that Rambus is suing over today was newly created by/at JEDEC. ALL OF IT can be traced back to the original Rambus patent application. You can, however, thank JEDEC for DELIBERATELY "inventing" an inferior memory (SDRAM) product to existing state-of-the-art RDRAM by only incorporating part of Rambus' amazing tech and then making YOU buy it!

Parent

RE: .

By Tanclearas on 4/27/2006 8:36:38 AM , Rating: 2

The trial judge who presided over the trial explained:

“[T]he evidence clearly and convincingly showed that Rambus committed actual fraud wantonly and maliciously and in total disregard of the rights of Infineon and all other JEDEC members by coupling its deliberate, pervasive, long-term violation of JEDEC’s disclosure policy with its implemented strategy to take advantage of its fraud by constantly using information (made known to it only because it was a JEDEC member) to modify its patent applications to assert claims which Rambus intended to cover the technology of the JEDEC standard.” (Order Denying Motion for JMOL, at pp. 73-74.)

Basically, there are people claiming we are using Rambus technology today because Rambus modified existing patent applications to incorporate what they were picking up at JEDEC.

As for the comment about inferior SDRAM vs state-of-the-art RDRAM, RDRAM was simply able to provide higher bandwidth, but did so at the expense of latency. The CPU's of the time had little use for the bandwidth offered by RDRAM, so choosing RDRAM over SDRAM would have meant a performance penalty across the board. By the time CPU's actually needed more bandwidth, Rambus was busy suing everybody associated with memory. It's no big surprise that the industry snubbed them for DDR SDRAM to get the bandwidth they needed.

Parent

RE: .

By dexvx on 4/25/2006 8:02:31 PM , Rating: 2

Impede progress?

DRAM manufacturers set us back at least a few years by pushing the inferior DDR SDRam over RDRam (and later XDR). There is no question that RDRam was superior to DDR SDRam at the time. And nowadays, there is no question that XDR and FlexIO is superior to DDR2/DDR3.

Parent

RE: .

By Tanclearas on 4/26/2006 10:24:29 AM , Rating: 2

Please tell me you're kidding. RDRAM offerred higher bandwidth at the expense of latency. The CPU's at that time had no use for the added bandwidth. The Pentium 4 could have used RDRAM bandwidth, but I think it's obvious that RDRAM would have hindered Athlon/Opteron/Athlon 64 performance. To say that "there is no question that RDRAM was superior to DDR" is laughable. RDRAM was suited to the Pentium 4, which has basically been the laughing stock of the CPU industry for the last few years.

Parent

Memory Prices

By BladeVenom on 4/25/2006 10:46:26 AM , Rating: 3

I guess this means memory prices are going up soon?

RE: Memory Prices

By dgingeri on 4/25/2006 10:56:22 AM , Rating: 2

yep. it also means that Rambus now has a strangle hold on our memory technologies as a whole, including DDR2. This is bad. very bad.

Parent

RE: Memory Prices

By kalaap on 4/25/2006 11:34:37 AM , Rating: 2

Not only the makers of DDR2 but also GDDRx and video card makers who use them.

Parent

RE: Memory Prices

By atridi on 4/26/2006 4:06:20 PM , Rating: 2

While that may be the "party line" from memory makers, its just not true. In fact, if the memory makers hadn't used price-fixing and deliberate manufacturing delays to kill RDRAM , they wouldn't have been able to artificially set high prices on DDR, and you would probably have paid less, at least in the past! That's why they paid the DOJ almost $1 billion in fines and sent almost a dozen of their collective execs to prison! Rambus isn't the enemy here, Hynix(Hyundae), Infineon(Siemens), Micron and Samsung are. if it weren't for Rambus tech, your computer would still have 60MHz EDO, remember that?

Parent

Rambus the theif

By Trisped on 4/25/2006 11:21:39 AM , Rating: 2

So Rambus comes up with a crappy new type of RAM. It fails because Rambus sucks as did their RAM idea. They sue everyone who makes RAM because they say they invented it first. The people offering patents need to stop handing them out over any little advancement. Most of the law suits now are over simple tech that I could have come up with in high school, like iTunes selling music on line which can be downloaded directly to a player.

RE: Rambus the theif

By josmala on 4/25/2006 12:55:41 PM , Rating: 3

Actually RAMBUS has made great RAM.
There where few things that made it fail.
First the manufacturers decided they didn't want it to succeed so they jointly decided to increase the price of DRDRAM.
2ndly first incarnation of chipsets using DRDRAM where crappy with weak FSB.
The 2nd generation DRDRAM chipsets with Pentium4 beat the hellout of any DDR products.

QUAD channel DRDRAM uses slighly more pins than SINGLE channel DDR and lot less than dual channel DDR. So basicly the argument on having latency with multiple modules is moot since you could make the DRDRAM interface per module, and still have more modules than equivalent unbufferent DDR. But intel chooce to go for cheaper chipset instead of higher performance in this matter.

Parent

RE: Rambus the theif

By Tegeril on 4/25/2006 11:37:55 PM , Rating: 2

I can attest to this beating the hell out of DDR. My 3GHz Northwood P4 (533fsb) is cheerily accompanied by 4x256MB PC1066 RDRAM (a perfect fsb increment). The SiSoft Sandra memory bench beat the shit out of DDR ram for over a year after 1066 stopped being made.

Parent

what's worse

By brownba on 4/25/2006 1:00:55 PM , Rating: 2

What's worse, companies constantly suing for money,
or all you guys ignorantly bashing them for doing so?

Some of comments on here are so ignorant...
I don't know the specifics of this case (and I doubt any of you do), but if you invented something, patented it, then a court found another company to infringe upon that patent, hell yeah any one of you would go after that company.
If not, then there's no reason for patents. (that's another discussion)

RE: what's worse

By SoylentG on 4/25/2006 1:24:01 PM , Rating: 2

Lets see...they're suing other companies for making ram...yet I haven't used that company's ram in many many years. I don't think I've even seen it on a shelf or an online store in recent memory.

Yet they get rediculous amounts of money from these companies which we use, all because they did something similar before. From what I've gleaned they just reinvented the wheel, and patented it, and are now suing many companies for doing something similar. Pfft.

Parent

RE: what's worse

By atridi on 4/26/2006 4:35:27 PM , Rating: 2

You, sir(or madam), have never STOPPED using their memory! The memory in the computer you type on today (assuming its even remotely current) runs as fast as it does because the signal lines are tied to (synchronous) an external clock signal, deskewed by on-chip delay-locked loop circuitry and triggered on both edges of the clock. In addition, response of the memory is controlled by an on-chip command register with programmable CAS delay. All innovations which did not/would not exist without Rambus. The memory you use today, sir, is definitely their design!

Parent

By Griswold on 4/25/2006 11:40:50 AM , Rating: 2

Thats Rambus' Q2 revenue. Well done, now start sueing... err.. working on Q3 revenue.

RE: Q2

By oringo on 4/25/2006 12:43:09 PM , Rating: 1

It doesn't matter if Rambus is an US company or not. Patterns like these hurt consumers eventually since the alleged pattern infringing producers now have to pay royalty to Rambus.

Parent

Sweet Revenge

By Commando on 4/25/2006 5:35:40 PM , Rating: 2

Rambus deserves there fair share after all they invested millions of dollars to create a technology we now use today.
A very good one i might add
Well done Jury .

Standards should explicitly state their IP costs

By peternelson on 4/26/2006 11:01:23 AM , Rating: 2

There is an easy solution.

When a standard or proposed standard is developed, a list should be made of all companies whose intellectual property it uses, what that property is, and the charges it will be made available for.

Then decisions can be made about whether it is a worthwhile standard to adopt.

I don't think it's good to come in years later and ask for money then.

Also in the interests of competition, several standards are useful, ideally based on different vendors rights.

I agree with the other poster too, if there was less arguing over who owed who what, then we could take the lawyers out the equation.

It's not just us paying for the RAM cost, but also the IP of the ram technology. Now that would be not so bad. As a consumer what I object to is paying toward millions of $ of legal fees.

what price non Rambus owned IP ram exists today ?

By david99 on 4/26/2006 10:24:33 PM , Rating: 2

what price non Rambus owned IP ram exists today ?

by that, im trying to get a feeling for, and if todays newest ram tech would be helped or hindered by this
US ruling and how it might effect the world markets and future speed increases.

what tech exists today that the world ram makers could use
and is it faster/better than current rambus IP?.

if above exists and is better, would it be more likely
that the world manufacturers would pay rambus their due
and continue to produce what already exists or would it
cheaper long term to move over to new tech assuming it exists today.