Nvidia Sues Qualcomm, Samsung Over GPU Patents

[piesquared]

There's a new patent troll in town I see. Nvidia probably doesn't see a very rosy future for itself... Patent trolling is something that's usually done in desperation. Alienating yourself from the industry isn't usually the best strategy lol. Could also be in nv's style of walling up the garden. Either way, terrible strategy. Utterly tragic decision and a fool's errand.

 

[Keysplayr]

There's a new patent troll in town I see. Nvidia probably doesn't see a very rosy future for itself... Patent trolling is something that's usually done in desperation. Alienating yourself from the industry isn't usually the best strategy lol. Could also be in nv's style of walling up the garden. Either way, terrible strategy. Utterly tragic decision and a fool's errand.

Nvidia's attorneys apparently don't agree with you.

Don't worry. It's not AMD they are going after. They have extensive cross-licensing between them.

You know, I just read that ParkerVision had sued Qualcomm in 2011 and Qualcomm was ordered by Jury to pay ParkerVision 173-million in patent infringement damages. Despite the conclusion the jury arrived at, a single federal judge overturned it. Basically voiding the 173million verdict.

This could happen again of course, but my point is, initially, Qualcomm was found guilty of patent infringement by a jury before it was overturned by a single judge. They may as well have not even gone to trial if this sort of thing can happen.

 

[caswow]

Nvidia's attorneys apparently don't agree with you.

Don't worry. It's not AMD they are going after. They have extensive cross-licensing between them.

You know, I just read that ParkerVision had sued Qualcomm in 2011 and Qualcomm was ordered by Jury to pay ParkerVision 173-million in patent infringement damages. Despite the conclusion the jury arrived at, a single federal judge overturned it. Basically voiding the 173million verdict.

This could happen again of course, but my point is, initially, Qualcomm was found guilty of patent infringement by a jury before it was overturned by a single judge. They may as well have not even gone to trial if this sort of thing can happen.

often the jury itself is the problem

 

[Keysplayr]

often the jury itself is the problem

Absolutely. Despite stringent jury selection approved by both sides. Especially when you want them to be made out to be the problem. Someone obviously did. My guess.

 

[krumme]

Its 17 years ago the "Rendering Pipeline" with its Screen space tiling was filed.

17 years and now we have a lawsuit.

Clearly this is a strategic move and have nothing to do with the - if - actual breaking of the patent.

 

[Keysplayr]

Its 17 years ago the "Rendering Pipeline" with its Screen space tiling was filed.

17 years and now we have a lawsuit.

Clearly this is a strategic move and have nothing to do with the - if - actual breaking of the patent.

Can you show us what aspect of the rendering pipeline is the concern? Or is it the entire "idea" of a rendering pipeline that is being called into question?

Clear? yeah. Clear as mud.

I'm not outright defending either. "'m probably reacting to the outright claim of wrong doing by Nvidia here before anybody knows anything about anything. No benefit of doubt given but plenty of blame for being in the wrong when most, if not every single person here cannot decipher "properly" what aspects of what IP is what Qualcomm is being sued over. Just stop.
The details WILL be provided eventually.

 

[krumme]

Can you show us what aspect of the rendering pipeline is the concern? Or is it the entire "idea" of a rendering pipeline that is being called into question?

Clear? yeah. Clear as mud.

I'm not outright defending either. "'m probably reacting to the outright claim of wrong doing by Nvidia here before anybody knows anything about anything. No benefit of doubt given but plenty of blame for being in the wrong when most, if not every single person here cannot decipher "properly" what aspects of what IP is what Qualcomm is being sued over. Just stop.
The details WILL be provided eventually.

Dont derail the point.

As you say its clear as mud what is in concern in the pipeline, as NV have not cared to explain it.

What is damn clear is this patent was filed 17 years ago. If this patent was broken, it has been for 10+ years. Why did it take 10 years for NV to react?

 

[Keysplayr]

Dont derail the point.

As you say its clear as mud what is in concern in the pipeline, as NV have not cared to explain it.

What is damn clear is this patent was filed 17 years ago. If this patent was broken, it has been for 10+ years. Why did it take 10 years for NV to react?

I'm not derailing a thing and don't accuse me of doing so.

What do you mean NV has not "cared" to explain it? To whom? the public? Meaning us? Cause you could probably be damn sure that it has been explained to Qualcamm and Samsung. Don't you think? Are we owed an explanation? Owed? No. Eventually it will be discussed publicly and in detail that goes beyond you and I. The stockholders are probably owed and explanation though. If you own stock then you should have received a stockholders notification or at least will, am I correct? I don't know because I have never owned any stock nor am I interested in doing so.

The "point" is not what you say it has to be anyway. 17 years or 17 minutes matters why? Is there a statute of limitations?

 

[Zodiark1593]

I don't see how this is any worse than Qualcomm getting licensing fees for nearly every modem sold, or Microsoft collecting on every Android sold. Given Nvidia's foothold in PC and HPC, I don't think they have much to lose being called morally bankrupt, and plenty to gain.

 

[krumme]

The "point" is not what you say it has to be anyway. 17 years or 17 minutes matters why? Is there a statute of limitations?

What is perfectly clear is, if the rights from this patent have been stolen, it have been stolen by most of the entire industry, and NV have known for 10+ years.

5 or 10 years ago it was not a problem going to court with. Now it is. What have changed since this is the case. As the tech and the patent is the same, it surely can not be about breaking the patent.

 

[krumme]

Samsung have show they go to extreme measures to steal technology eg. Oled tech from LG partners. They dont give a rats ass about anyting but money, and treat their employees as robots. They copy whatever they can and do it all the time. But when NV is suing now, it can only be because of the entire Tegra strategy is going to a halt. They can not sell their mobile solutions, and the next step is therefore to change strategy and sell the IP.

 

[Phynaz]

Too many people don't know the definition of patent troll.

 

[Anarchist420]

nvidia isnt a patent troll (they may or may not have been helped by the patent system, but they definitely have shown the world they are apt at marketing). however, this suit is not going to help nvidia long term if they continue to actively support the system. i dont know why they dont surrender their patents since they have good engineers and since JH has shown himself to be business savvy and very efficient.

samsung is a walking timebomb, and there will be more bad signals (patents trick people first) if they lose a patent lawsuit. maxwell doesnt sound like it is anything all that innovative. but if they can get their drivers straight for all maxwell users, then things may appear better than what they actually are. drivers for fermi and kepler have not been good and weak integer performance coupled with crippled double float performance was not good for replication of old features some which were calculated with 32 bit integer precision. i dont know why they dont promote double float performance and fast lossless conversion of ints to double float for all of the products they make.

that said, less inside the box action by a corporation is generally better than more inside the box action and suing some poor dude like samsung for infringing on patents helps the State more than it could ever help any individual.

 

[Mopetar]

Mopetar, the infringed patents are listed in the register article:
http://www.theregister.co.uk/2014/09/04/nvidia_sues_qualcomm_and_samsung/

Thanks. They also have a list of them, as well as some other interesting details in filed complaint.

 

[Keysplayr]

Samsung have show they go to extreme measures to steal technology eg. Oled tech from LG partners. They dont give a rats ass about anyting but money, and treat their employees as robots. They copy whatever they can and do it all the time. But when NV is suing now, it can only be because of the entire Tegra strategy is going to a halt. They can not sell their mobile solutions, and the next step is therefore to change strategy and sell the IP.

Wow. You know so much that you don't know anything. Only ones that do are Nvidia, Qualcomm and Samsung.

 

[Abwx]

Problem is they patented it:
https://www.google.com/patents/US6198488

What? They did it a a time where no other company was able to do it. ImgTech wasnt able to it with Kyro 2...

Patent was deposited :

6 déc. 1999

Let s see what did exist at the time :

During 1999, Diamond and S3 merged and the Savage 2000 GPU was the first product from the combined companies


Savage 2000 supported S3's S3TC texture compression, a hardware transform and lighting engine named "S3TL", and was equipped with a "QuadTexture Engine" capable of a single quad-textured pixel per clock or 2 dual-textured pixels per clock.

At the end of 1999, the Savage 2000 and the GeForce 256 were the only gaming-oriented cards with hardware T&L.

A troll patent by the definition since there was prior and applied art, that S3 got his TL being bugged is another matter, Via who bought S3 could as well claim intellectual property being infringed and ask for some money....from Nvidia.

 

[sontin]

Despite the earlier launch date of the Geforce SDR the hardware T&L unit of the Savage 2000 was not working at launch and AFAIK they never released any D3D7 drivers for it.

Diamond eventually released drivers with S3TL support in OpenGL and Direct3D. Unfortunately S3TL does not function properly. It causes missing textures, errors in geometry and models, and minimal performance benefits. Whether the issues are a result of poor drivers or defective hardware is unknown.

http://en.wikipedia.org/wiki/S3_Savage#Savage_2000

Prior Art, sure.

 

[monstercameron]

Despite the earlier launch date of the Geforce SDR the hardware T&L unit of the Savage 2000 was not working at launch and AFAIK they never released any D3D7 drivers for it.


http://en.wikipedia.org/wiki/S3_Savage#Savage_2000

Prior Art, sure.

lol all that build up to try and spin it...and you just gave up. I have to admit I am a bit dissapointed. :$

 

[Mopetar]

A troll patent by the definition since there was prior and applied art, that S3 got his TL being bugged is another matter, Via who bought S3 could as well claim intellectual property being infringed and ask for some money....from Nvidia.

Just because someone else did something similar doesn't mean that nVidia's implementation isn't patentable or that others aren't infringing on it. You would need to show that the implementation created by S3 is exactly the same as described by nVidia's patent.

Otherwise it doesn't necessarily qualify as prior art. There were several different designs for engines for the rotary engine was invented. Would you claim that it wouldn't be worthy of a patent just because someone else had already built an engine?

There are too many people who have almost no understanding of how patents or patent law actually works trying to throw around their opinions as if they mean something.

 

[96Firebird]

There are too many people who have almost no understanding of how patents or patent law actually works trying to throw around their opinions as if they mean something.

Yup, plenty of armchair lawyers in this thread. Good for a laugh though... ^_^

 

[Exophase]

Interestingly, nVidia isn't being honest when they say this is the first time they've filed a lawsuit against another company regarding their GPU patents. They did it to 3dfx in 2000, before acquiring them:

http://techreport.com/news/1131/nvidia-clubs-3dfx-with-patent-suit

Look how similar the defense statement is to the one they issued recently..

These patents look quite a bit more specific than many of the ones in this current case.

Just because someone else did something similar doesn't mean that nVidia's implementation isn't patentable or that others aren't infringing on it. You would need to show that the implementation created by S3 is exactly the same as described by nVidia's patent.

How is it that you seriously think that anyone else is using a T&L unit like the one in GeForce 256 in a modern mobile GPU? No one has been using fixed function T&L for several years to begin with. Your argument is self defeating.

 

[Keysplayr]

Interestingly, nVidia isn't being honest when they say this is the first time they've filed a lawsuit against another company regarding their GPU patents. They did it to 3dfx in 2000, before acquiring them:

http://techreport.com/news/1131/nvidia-clubs-3dfx-with-patent-suit

Look how similar the defense statement is to the one they issued recently..

These patents look quite a bit more specific than many of the ones in this current case.



How is it that you seriously think that anyone else is using a T&L unit like the one in GeForce 256 in a modern mobile GPU? No one has been using fixed function T&L for several years to begin with. Your argument is self defeating.

I wouldn't take any arguments generated in here, for or against, (including yours and mine) what the details of the suit are and what exactly the IP is that is being allegedly infringed upon. You have no idea what you're looking for in the patents. Then or now. Nor do any of us. Appreciate that.

 

[Zanovar]

hahah,that post i so funny keys.

 

[Zanovar]

unintentionally im guessing

 

[Exophase]

I wouldn't take any arguments generated in here, for or against, (including yours and mine) what the details of the suit are and what exactly the IP is that is being allegedly infringed upon. You have no idea what you're looking for in the patents. Then or now. Nor do any of us. Appreciate that.

This whole line of reasoning that the patents is written in magic code and could contain anything is a little overblown, don't you think? nVidia has enumerated which claims in which patents they think are being infringed. Now of course I can't predict what will happen in the court but it's pretty obnoxious to say that I'm not allowed to read the claims and have an opinion on their validity and applicability.