Two Rambus Patents Invalidated By USPTO

First time accepted submitter rnswebx writes "Two patents that chip designer Rambus used to win patent lawsuits against Nvidia, HP, and others have been declared invalid by the USPTO." The Inquirer has a similar story up, with appropriately snarky sub-head.

Comments

Progress?

[Anonymous Coward]

Well, two steps forward? how many back?

Patent question

[j00r0m4nc3r]

If a company is sued by a patent holder and forced to pay licensing fees, and that patent is later invalidated, is that company entitled to reimbursement?

Re:Patent question

[gstoddart]

If a company is sued by a patent holder and forced to pay licensing fees, and that patent is later invalidated, is that company entitled to reimbursement?

Likely not. They'd likely argue that at the time they held a valid patent, and that you entered into a licensing agreement with them.

So, stupid people can issue lousy patents ... but when that patent gets overturned, anybody who got extorted to pay for it is likely out of luck.

Re:Patent question

[AikonMGB]

Inquiring minds want to know. If the use of said patents was the sole factor in the suit, then I would say, the company is entitled to reimbursement. If there was anything else going on at the time, or if for example it involved two patents and only one patent was invalidated, then I think the case would have to be revisited. IANAPL, this is just wild speculation.

What I want to know, however, is how those patents made it through said suits intact, only to be declared invalid by the USPTO at a later date -- wouldn't, in the course of the suits, the patents get thoroughly vetted by the USPTO, under scrutiny of the court? Or am I applying too much logic and common sense to the patent system?

Aikon-

Re:Patent question

[gstoddart]

wouldn't, in the course of the suits, the patents get thoroughly vetted by the USPTO, under scrutiny of the court

Nope. The patent stands until the USPTO invalidates it. They're motivated to churn through patents and get paid for them ... the USPTO isn't necessarily interested in making sure the patents are any good. And, the legal standard to overturn a patent is quite high. (Sorry, it's only a partial article)

Or am I applying too much logic and common sense to the patent system?

Far too much, sadly. It's a rigged game, with bad outcomes.

Re:Patent question

[Grand Facade]

If anyone is liable it it the patent office as the judgements were granted based on their decision.

Good luck collecting on that....

Re:Patent question

[Defenestrar]

Don't forget cases in which a company (e.g. NVIDIA) settled with Rambus.

Previously, the International Trade Commission (ITC) ruled that Nvidia, HP and other firms were infringing two Barth I patents owned by Rambus. This led to Nvidia settling with Rambus.

The courts ruled that there was an infringement, but the deal to use the technology after that wasn't court mediated. That could mean a reversal in the patent status won't automatically even give valid justification for a refund - it'll depend on the settlement terms. If it's written such that one party is paying for technical expertise, design parameters, surcharge per unit installed, or etc... it could be pretty murky.

For example, if someone promised to give me a penny for every breath they took, even if I don't own their air, I should still be able to collect on that (of course it's again murky if I had previously convinced them I owned the air, or I would poison it if they didn't pay, ...). Probably the most common version of this sort of agreement that I run across is when some kid comes up to me and asks for $0.25 to go toward cancer research fundraiser personnel for every lap he walks on the high school track.

Re:Patent question

[AikonMGB]

The courts ruled that there was an infringement, but the deal to use the technology after that wasn't court mediated. That could mean a reversal in the patent status won't automatically even give valid justification for a refund - it'll depend on the settlement terms. If it's written such that one party is paying for technical expertise, design parameters, surcharge per unit installed, or etc... it could be pretty murky.

Hmm, a good point; I've never liked the concept of settling from an idealistic point of view, but this would certainly be an instance that actually backs that up.

Aikon-

Re:Patent question

[Anonymous Coward]

This is, in fact, part of "The Problem".

The courts, by default, assume that the USPTO did their job correctly and **all** issued patents are valid.
Someone who infringes on an obviously invalid patent needs to pay the cost of delaying their trial while they, on the side, try to get it invalidated by the USPTO. Judges who "get it" will regularly stay the trial pending the results of the invalidation examination, which helps a little, but even that isn't guaranteed.

Re:Patent question

[Yobgod Ababua]

This is, in fact, part of "The Problem".

The courts, by default, assume that the USPTO did their job correctly and **all** issued patents are valid.
Someone who infringes on an obviously invalid patent needs to pay the cost of delaying their trial while they, on the side, try to get it invalidated by the USPTO. Judges who "get it" will regularly stay the trial pending the results of the invalidation examination, which helps a little, but even that isn't guaranteed.

[repost with correct login]

Re:Patent question

[geekmux]

If a company is sued by a patent holder and forced to pay licensing fees, and that patent is later invalidated, is that company entitled to reimbursement?

If a company creates a highly specialized type of memory module, only to have it crushed into early obsolescence 6 months later, am I entitled to a reimbursement for this worthless RAMBUS motherboard?

Re:Patent question

[tlhIngan]

If a company is sued by a patent holder and forced to pay licensing fees, and that patent is later invalidated, is that company entitled to reimbursement?

It depends. If they reach a settlement, then the terms of the settlement determine what happens if the patents are found invalid in the future. Since the terms of such settlements are often confidential, we'll never know.

For court-ordered settlements, most likely not, because if the patents were contested, the court will wait to see if the USPTO revokes those patents. If they're later overturned, well, it depends how good the lawyers are - perhaps they can wiggle some way of suing for that money back (the appeal avenue would've been lost).

Now, if a company is paying license fees, it too also depends on the licensing agreement - some will offer refunds (not a popular option, but depends on the customer), others partial, and of course, many none. A justification of that is that those who paid had a temporary advantage over those who didn't, so the license fees paid would've been made up by time-to-market.

Re:Patent question

[Saxophonist]

For example, if someone promised to give me a penny for every breath they took, even if I don't own their air, I should still be able to collect on that (of course it's again murky if I had previously convinced them I owned the air, or I would poison it if they didn't pay, ...). Probably the most common version of this sort of agreement that I run across is when some kid comes up to me and asks for $0.25 to go toward cancer research fundraiser personnel for every lap he walks on the high school track.

I am not a lawyer, but one characteristic of a valid contract is consideration. Each party to a contract must incur a legal detriment of some kind -- that is, each kind needs to give up something of value, and each party needs to receive something of value. I don't know that either example you give constitutes a valid contract because neither has consideration. Also, the first agreement was probably made by someone not of sound mind, and the second was made by a minor.

Pedantic enough yet?

As it applies to patents, though, if party A licenses a patent from party B, and the patent is later found to be invalid, and somehow no lawyer on either side thought to include any language on this situation (which means the parties should really find better lawyers), my guess is that the contract would be invalid due to lack of consideration. Of course, that's only a guess. If there are ongoing licensing fees, though, I would imagine party B would have a tough time collecting on any future fees.

Any lawyers are encouraged to correct what I have said here, as it is probably dangerously misinformative in some way.

Re:Patent question

[inviolet]

Nope. The patent stands until the USPTO invalidates it. They're motivated to churn through patents and get paid for them ... the USPTO isn't necessarily interested in making sure the patents are any good.

Indeed. At more than $1000 filing fee per patent, the USPTO is now a revenue center for the Federal Government. We lack the political will to raise taxes or to lower handouts, so we resort to sneaky, lossy, eventually ruinous methods of revenue. Like patent applications.

So yeah, the incentives at the USPTO are completely perverse.

Re:Patent question

[Theaetetus]

Inquiring minds want to know. If the use of said patents was the sole factor in the suit, then I would say, the company is entitled to reimbursement. If there was anything else going on at the time, or if for example it involved two patents and only one patent was invalidated, then I think the case would have to be revisited. IANAPL, this is just wild speculation.

License agreements are contracts, so you'd have to look to the specific terms in the contract. Most will include provisions that terminate the obligation of royalty payments if the patents are invalidated, but very few require repayment of past license fees - because the patent owner would never agree to such provisions.

What I want to know, however, is how those patents made it through said suits intact, only to be declared invalid by the USPTO at a later date -- wouldn't, in the course of the suits, the patents get thoroughly vetted by the USPTO, under scrutiny of the court? Or am I applying too much logic and common sense to the patent system?

Most of this stuff doesn't have bright line tests, but instead uses questions of fact that could get decided either way - and that's why these things go to court. If they were always certain, there would never be lawsuits. One jury may find the patent valid, while another jury (or, in this case, the Examiner) looking at the same evidence may find the patent invalid. Or maybe the second jury/Examiner is looking at new evidence. And this isn't final - the Examiner could still be reversed by the court.

Re:Patent question

[Dragonslicer]

wouldn't, in the course of the suits, the patents get thoroughly vetted by the USPTO, under scrutiny of the court

Nope. The patent stands until the USPTO invalidates it.

You're both incorrect.

The USPTO is not directly involved in litigation, so the patent is only vetted by the court as part of the lawsuit. The USPTO can also reexamine a previously issued patent, but this is generally independent of a specific lawsuit.

A federal court can invalidate a patent on its own as part of a lawsuit. The USPTO is not directly involved with this decision.

Re:Patent question

[jbengt]

What I want to know, however, is how those patents made it through said suits intact, only to be declared invalid by the USPTO at a later date

IIRC, there was a motion to delay the court proceedings until the USPTO ruled, but the judge ruled that the court case couldn't wait for the USPTO review.

Re:Patent question

[AikonMGB]

IIRC, there was a motion to delay the court proceedings until the USPTO ruled, but the judge ruled that the court case couldn't wait for the USPTO review.

Wow, what a gong show =/

Aikon-

Re:Patent question

[Anonymous Coward]

Yeah, right! Patents are a crock.

Re:Patent question

[Comrade Ogilvy]

That is a point of negotiation.

A licensse can ask for a lower price in return for a provision that the ultimate validity of the patent is irrelevant. Or the opposite. Obviously, the implicit threat is a drawn out and uncertain legal battle where the would be licenser may end up with nothing buy legal fees.

Re:Patent question

[sjames]

If it later turns out the kid didn't actually give the money to the charity or that the "charity" was actually "buy a kid a new xbox", he (or his parents) are on the hook to either refund the money or make good on the donation to the designated charity.

Re:Patent question

[micheas]

The court presumes that the patent is valid, so unless there is clear and convincing evidence then the patent stands.

The courts presume that all patents granted are valid and that it is extremely unlikely that an invalid patent is before them. This is of course most likely wrong, and the cause of much of the nonsense in patent cases.

One could probably move to revisit the case, but I am not familiar with the rules of court in federal courts.

Re:Patent question

[Errol backfiring]

They'd likely argue that at the time they held a valid patent

No. It is the same patent. If it is invalidated, it is because it is an invalid patent. It is not a patent that rotted, wore out or turned sour. It is the very same as it always was. An invalidation is a correction to the fact that the patent was granted at all. Off course, the patent office should be held responsible for the damages.

Barth I

[Anonymous Coward]

It appears that it deals with 2 out of 3 patents, collectively known as Barth I.

If you are sentenced to death by hanging, quartering and finally beheading, and another judge orders you cannot be hanged or quartered, you may win, but you will still lose your head (perhaps we were better off with car analogies...). In other words, the third patent still stands.

money back?

[Kvasio]

So ... do Nvidia and HP get their money back? How does US legal system solve such situations?

Re:money back?

[Alex Belits]

It doesn't.

It doesn't even investigate Rambus for fraud.

Re:money back?

[drinkypoo]

If someone has a smoking gun that proves that someone making decisions for RAMBUS knew that they were making specious claims or whatever grounds the patents were invalidated upon then the injured parties probably have a case for civil court. But even though everyone knows that's what happened, the courts are not likely to even hear the case without some proof.

Re:money back?

[Anonymous Coward]

"How does US legal system solve such situations?"

By saying "Hey, it's not my problem."

Re:money back?

[sjames]

By blowing a raspberry, thumbing it nose and saying "Not my problem! So long, suckers!"

To paraphrase, of course.

I remember when Rambus made RAM

[NixieBunny]

They were exotic fast things, and they didn't always behave properly, but at least they were actual products. Isn't that the ostensible reason why companies exist, to make products?

Re:I remember when Rambus made RAM

[gstoddart]

They were exotic fast things, and they didn't always behave properly, but at least they were actual products. Isn't that the ostensible reason why companies exist, to make products?

No, to make money is the reason companies exist. The legal system makes no requirements as to how you make your money.

As long as patents are something you can buy and sell, there will always be companies who own them just to squeeze licensing fees out of companies who actually do make things.

Welcome to capitalism, it's supposed to be the best thing ever.

(And, really, RAMBUS basically sat on standards committees and then filed submarine patents on the very stuff they and others had designed and agreed upon ... seeing their patents get invalidated warms my heart. They're a bunch of bottom feeders.)

Re:I remember when Rambus made RAM

[Anonymous Coward]

Uhm... they never made RAM. The licensed their patents to other companies that actually made memory modules. Those modules were just referred to as Rambus memory.

Re:I remember when Rambus made RAM

[compro01]

They never made RAM, they came up with the specification for RDRAM. And they then got their lunch eaten when DDR became available for the Pentium 4 and offered almost as good performance for much cheaper and without the eccentricities and then went sue happy over it.

Though their new XDR2 RAM looks promising. It's supposed to be used in AMD's upcoming 7900-series videocards.

Re:I remember when Rambus made RAM

[geekmux]

They were exotic fast things, and they didn't always behave properly, but at least they were actual products. Isn't that the ostensible reason why companies exist, to make products?

Er, make products? If by "make" you mean become a Patent Pimp, whoring out licensing and suing anyone who gets within ten feet, then yes, that is the intent of business today.

Re:I remember when Rambus made RAM

[Anonymous Coward]

Course, the other RAM manufacturers back then engaged in price-fixing in order to stifle competition from Rambus...

Re:I remember when Rambus made RAM

[KingMotley]

Actually, DDR was never as fast as RDRAM, but it had better latencies. RDRAM was technically a better solution at the time, and becomes even more so as we scale up in processor frequency and number of cores, but the royalties (and memory manufacturers who colluded to ruin RDRAM through price fixing) killed it.

Re:I remember when Rambus made RAM

[tlhIngan]

They were exotic fast things, and they didn't always behave properly, but at least they were actual products. Isn't that the ostensible reason why companies exist, to make products?

While they don't make the chips themselves, they still do sell RDRAM.

In fact, it's been used in at least two generations of popular consoles. The PS2 had 32MiB of RDRAM in it. The PS3 has 256MiB of XDRDRAM (and 256MiB of GDDR3, I think).

Sony's gone RAMBUS for over a decade. The licensing fees alone for the PS2 and PS3 probably amount to a good chunk of their income.

Re:I remember when Rambus made RAM

[nschubach]

I'm not sure how Patents and Capitalism go hand in hand... in fact, I'd think that Patents inhibit Capitalism.

Re:I remember when Rambus made RAM

[Agripa]

That whole fiasco with not being able to use all of the RDRAM slots on a motherboard because of poor reliability did not help them any.

I avoided them just because of their sleazy business practices. I did not care if what they did was legal or not.

Re:I remember when Rambus made RAM

[sjames]

They were exotic, but turned out not to be so fast for real world access patterns.

HP and nV must feel great

[bhcompy]

They had to pay Rambus a bunch of money to give them a rimmjob they didn't have to. And I bet Rambus didn't take a shower before making them do that.

Re:HP and nV must feel great

[Kvasio]

judging by their recent activity, HP still did not get over that fart

USPTO

[DinDaddy]

My first thought upon reading this was "They should make Rambus repay the companies that gave them licensing fees based on these." My second was "with treble damages to discourage this shit." My third was "On the other hand, they were just (underhandedly) playing the system as it exists so maybe that's a little unfair". Then I realized the solution. Make the USPTO repay the license fees. That would improve the quality of their patent review really really fast.

In my dreams. It would more likely mean no patent ever got overturned.

Re:USPTO

[medv4380]

No that would slow down their patent review even more. If we just invalidated all software and method patents their work load would be reduced and would be able to do a better job reviewing and hearing challenges to existing patents in a more timely manner.

Re:USPTO

[Anonymous Coward]

...or granted in the first place.

It could be argued the USPTO is negligent, and thus owes some relief to those left out of pocket, since:

a) there were some $ losses due to matters arising from the incorrect award of the patent
b) they knew (or should have known) incorrect award of a patent would result in such losses.
c) that incorrect award was of USPTO's doing
d) that a $ value can be put on the losses.

It simply isnt consistent with the idea of natural justice that:
a) BigCorp with $$$ applies for patent
b) USPTO grants patent
c) BigCorp sues LittleCorp
d) LittleCorp can either license for $ or fight (and cause re-examination) also for $$

Most LittleCorps would prefer to spend $ rather and $$.

The analogy often presented of The Mob shaking down a local small business is a good one.
However, in this case the USPTO is enabling the bad behaviour. They needto quit doing that.
And in order to quit they either need to modify their behaviour, which means sanctions against the USPTO and/or their employees when they screw up.

I refer you to what happens to hospitals/physicians when they hurt somebody.

Re:USPTO

[MimeticLie]

Wow, you've looked at a bad situation and found a way to make it even worse. Why would the USPTO invalidate any patents under your system?

Re:USPTO

[Richard_at_work]

Why would the USPTO issue any patents under his system, more like. A much better situation!

Make the USPTO pay for any mistakes they make in issuing that later costs either the receiving company money (in litigation costs they incur thinking they have a valid patent) or any company involved in litigation defence or licensing.

Re:USPTO

[Amouth]

A judge should also be able to rule that one is invalid.. if only the USPTO can make one invalid then you lose your check & balance.

Re:USPTO

[Java Pimp]

http://en.wikipedia.org/wiki/Treble_damages

Re:USPTO

[magarity]

Then I realized the solution. Make the USPTO repay the license fees.

??? The USTPO charges all its extra expenses not covered by fees to the taxpayers. Your idea == bad.

Re:USPTO

[denis-The-menace]

But that would also kill USPTO's main profit center.

Re:USPTO

[Theaetetus]

No that would slow down their patent review even more. If we just invalidated all software and method patents their work load would be reduced and would be able to do a better job reviewing and hearing challenges to existing patents in a more timely manner.

Hell, why stop there? Let's invalidate all composition of matter patents and article of manufacture patents. I mean, as long as we're arbitrarily ruling out an industry from access to patents, let's go whole hog.

Re:USPTO

[Anonymous Coward]

We should allow people to patent any mathematical equation!

Re:USPTO

[sjames]

As long as we make filing start the expiry countdown, it's fine. With any luck, all patents will already be expired by the date they're granted and corporations will lose interest in them.

Re:USPTO

[WNight]

Right. If we gave such a huge boost to one industry we'd have to do it for the rest.

Why?

[Splab]

I know it's slightly off for a slashdot reader, but I did read the articles, and none of them talk about why the patents where rejected.

They can apeal the rejection, so nothing is settled yet - but it would be nice to know on what grounds those patents where rejected; prior art? obvious or bribes?

Re:Why?

[condition-label-red]

...or insufficient bribes... :-)

Screw the money...

[xded]

What I'm asking now is: when will Nvidia and HP be able to use these patents content to improve their products? Did these patents cover something technologically significant?

Is an USPTO decision, for once, going to foster innovation?

Re:Screw the money...

[somersault]

If they lost, it doesn't mean that they couldn't use the technology - it just means that they have to pay a license to use it.

Re:Screw the money...

[mikael]

This battle has ranged on since 2002. It relates to memory controllers, decode logic and high-speed memory chips for motherboards and GPU's.

Wikipedia entry on RAMBUS is the best explanation.

Re:Screw the money...

[Reverand Dave]

Correct, I worked for a company that RAMBUS sued after it pulled it collaborative patents derived from the JEDEC membership and standards. Rambus then proceeded to sure everyone that made any kind of RAM for a patent they claimed to own, but was actually created using joint resources of the JEDEC committee. RAMBUS are the original patent trolls of the modern age.

Re:Screw the money...

[mikael]

Ah, "submarine patents". Reassure everyone that the technology you offer and share is free, then when the patent is granted, announce that there is now a licence fee.

The hope is that after four years, everyone will have embedded your technology into their products, and can't go back.

Re:Screw the money...

[Reverand Dave]

It pretty much lead to a global patent suit with rambus taking on every maker of DRAM on the planet once they dropped the shitty rambus memory standard and switched to DDR.

Of course

[Anonymous Coward]

Under a fair and just system of law, of course they would. That isn't even a question. The question is whether we live under a fair and just system of law, and I think we all know the answer to that.

And In Other News

[MightyMartian]

And in other news, the barn door was closed after the cows had come home.

patent litigation

[Anonymous Coward]

What happens next depends on the timing.

If a trial court has rendered a verdict, yet such has not become final (because of post-trial motions), the losing side in a patent case would file a motion for a reconsideration. If the trail verdict has become final (an entry of judgment), the next step would be to file an appeal. In patent cases, this is done to the Federal Appeals Court of the geographical area (the Federal District Court) which rendered the verdict (this is per Federal Statute). The statute for filing a Notice of Appeal is 30 days after the entry of judgment.

Now if the time for appeal has passed, the afflicted parties would file a new lawsuit, asking for relief from the previous judgment based on the fact that the patent was invalidated.

WHEN PATENTS ATTACK!

[OnTheEdge]

If you haven't listen to this podcast from This American Life yet, you should. http://www.thisamericanlife.org/radio-archives/episode/441/when-patents-attack

from the does-this-mean-hp-gets-a-refund? dept.

[Anonymous Coward]

n/t

Two possibilities

[voss]

1) If there is language in the licensing agreement to that effect.

2) If they can make the argument to a court that having to pay license fees for invalid patents is somehow unjust enrichment(especially
if the license fees are still being paid).
http://en.wikipedia.org/wiki/Unjust_enrichment

Sue the guilty party

[Citizen of Earth]

Everyone victimized by these ridiculous patents should get into a class-action lawsuit the guilty party in all of this -- the patent office. Adding punitive damages for the results of its fraudulence and incompetence, the lawsuit should run into the hundreds of quadrillions of dollars. Perhaps this would bring some needed sensibility to the process, like peer review or simply throwing it all out the airlock.

What I want to know is...

[Anonymous Coward]

How old were these patents in the first place? What a waste of time and money.