Apple may face $900m bill after A7 CPU in iPhones, iPads ripped off university's patent Apple's A7 processors in iPhones and iPads infringe a patent held by the University of Wisconsin, a court has ruled. A jury, sitting in a US district court in Madison, Wisconsin, found Apple guilty of patent infringement late yesterday afternoon. While the damages in the case have yet to be determined, Judge William Conley …
Creating verilog code is expensive. Proving that some verilog does what you intended is really expensive. I agree that Apple should pay for all the verilog code they stole from the patent. A quick look at the patent shows zero lines of verilog and zero evidence the the verilog does what the patent describes.
There is some theory that one among the thousands of people who independently came up with the idea should be able get a monopoly on its implementation. If something is so obvious that no-one writes it in a technical journal, it is proof (to a patent lawyer) that the idea is unique and valuable. This dates back to the Dewar bottle.
A Dewar bottle is a delicate piece of lab equipment that can hold hot or cold liquids and keep their temperature roughly constant. A judge upheld the patent on a thermos flask (a Dewar flask robust enough to survive getting dropped) because the idea of making a Dewar flask robust would not have occurred to anyone but the patent holder. The judge pointed at a picture of a Dewar flask to show how different it was from a thermos. In the background of the picture you can see the thermos made by the lab-tech who took it on pic-nics to the delight and amazement of his family. The lab-tech, James Dewar and hundreds of other lab-tech and physicists knew that the thermos was to obvious to patent. The key quality that is required is the willful ignorance of a patent lawyer.
On another day, I would happily slag Apple for suing others who build things that Apple copied, but not for this.
There is some theory that one among the thousands of people who independently came up with the idea should be able get a monopoly on its implementation. If something is so obvious that no-one writes it in a technical journal, it is proof (to a patent lawyer) that the idea is unique and valuable.
Many things seem obvious after the fact. In 1996, nobody had thought of it.
I think you may misunderstand the way engineering works. It is normally small steps, but the fact that it is a small step does not mean it is obvious.
Your example of the Thermos flask actually proves my point. If it was so obvious, why had nobody done it before? Just because you can look at it now and say, "Oh, that's easy, just take this piece of equipment, add this to it, and you've got something much more useful outside the lab" does not mean it should not be patentable. As long as there wasn't another lab tech out there who independently did the same before him, it's a perfectly valid patent.
Also, the theory is not that one among those who independently has an idea gets a monopoly. It is that the first person to come up with the idea gets a monopoly in exchange for publishing his idea.
EDIT: Forgot to add, as the idea was patented, it was also published. It probably can't be proven, but how do we know that an Apple engineer had not read the patent, or about it in some tech publication, or heard about it in discussions at University?
I'm sure there should be some distinction between innovation and application. In the example, the Dewar flask is the innovation, but deciding to keep coffee in it is a particular application. If I chose to deliver ice-cream in it, have I innovated?
Of course the subtleties of patent law make such trivial examples somewhat moot, but I'd argue the goal of patents should be the minimum legal structure possible to support investment in innovation - and no more. We can all agree that innovation is good, but artificial monopolies are bad.
Apple are a classic example of BigCorp. Are we actually surprised that they are without shame? Any company in the Western World would do what they are doing. The fact they they are fucking hypocrites after have sued Samsung largely for "look and feel" type patent breaches is neither here nor there. The BigBoys operate on what they can get away with. The only question they ask their legal departments is "can anyone what successfully sue/prosecute us". They are, none of them, interested in ethical conduct. The system stinks.
So the university sued Apple in Wisconsin since the Uni is located in Wisconsin. The patent troll route would have been to choose East Texas for the lawsuit venue.
Serving the papers in Hawaii would have probably delighted the Uni lawyers but would have been quite an expensive endeavour.
This post has been deleted by its author
"An American court ruling for an American company"
quel surprise
Since both the plaintiff and defendant were American companies, the real trick would have been for the American court to rule for a non-participating Congolese firm. Now THAT would have been 'quel surprise'.
Why apple and not arm ? Or have they really made large changes to the cpu ? Though in other articles it mentions apple quoted the patent in one of their patent so they must have realised something like this might happen ?
The initial intel case is pretty weird, intel appear to have paid for work and had the professors work with them only to find out they didn't get any patents ?
I dont know, it's not clear. However, from the linked document:
"9. The Court has personal jurisdiction over Apple pursuant to Wis. Stat. § 801.05(1)
because Apple is engaged in substantial and not isolated activities in this state and judicial
district, including maintaining a retail store and employees here. "
That doesnt apply to ARM
"16. Since the issuance of the ’752 patent, Apple has filed one or more patent
applications that cite the ’752 patent as relevant prior art. "
I don't know if ARM has cited patent '752
The patent is for a specific feature that optimizes load/store behavior that was not included in ARM's licensed 32 bit designs. I don't know for sure it isn't included in their 64 bit designs like the A53/A57/A72, but I doubt it. Since Apple doesn't use those but designs its own cores, it could be liable for something that doesn't affect those using ARM designs. It could affect Qualcomm or Samsung's own designs though.
They already settled with Intel for an unspecified amount, since Core i CPUs use this optimization as well.
AFAIK, Apple's A7 is based on ARM processors. To what extent does this decision affect other users of ARM processors? How do later Apple processors avoid infringement?
I guess I'll have to read about this decision more widely. Hopefully, the article will be updated later to better cover the implications.
See my reply above. This is over a specific implementation feature so other users of ARM are only affected if the specific implementation uses the same optimization. Apple may have the only ARM CPUs using this. They can only avoid infringement by not using this optimization.
The patent is actually fairly simple, to put it in a more real-world example it would be like if I told a cook "assume all men order their burgers with bacon, and all women order their burgers without, so make them that way when you see them walk in to save time - but if you see a man order one without or a woman order one with, keep track of that person so you don't make the wrong burger if they come back and waste time having to remake it". It may not be rounded corners, but it isn't exactly some huge innovation worthy of $862 million. But then few things granted a patent really are much of an innovation.
Oops - meant to include a note about ARM: Apple looked at ARM's 64-bit ARMv8 instruction set and designed a chip from scratch compatible with it, then later licensed the instruction set.
None of this is ARM's problem: it was Apple's custom system-on-chip design that infringed the uni's patent.
I've slung a sentence or two about this on the end of the article.
C.
Thanks for clearing that up.
Though, the facts as you state them make a mockery of the incessant fandroid blather here about Apple not designing chips and not having any chip design skills. Your suggestion is that aside from the instruction set, there is not much ARM in there at all?
Apple are now in fact one of the largest employers of chip designers in the world. Larger than all save Intel. Their ARM core design is Industry leading and is now providing them with a key advantage over the competition. They have been tailoring their chip design to perfectly suit iOS mobile usage.
"they have effectively surpassed the performance of Intel. This has been well documented."
http://stevecheney.com/on-apples-incredible-platform-advantage/
And now they are moving on to designing their own wireless chipsets.
Re: DEC. IIRC the Alpha had a lot of patented stuff and INTEL settled out of court to make the issue go away. The total was 3/4 Billion if memory serves where INTEL bought a DEC FAB as part of the settlement.
That said, it seems most likely that ARM has implemented the OoO capability, and if they are licensed for this patent, then Apple will be covered under the exhaustion principle. It all seems a little strange.
Apple had countered that the patent was invalid and had suggested WARF was a patent troll.
Seems to me that this will likely hold up in an outside WI court. If they had a previous private settlement that was paid out it looks pretty cut and dried. Just more Apple getting bent because they got caught stealing and are getting taken to task on it. Chumps.
Besides the mindless dronage of the typical iWhatever user, it is this type of arrogant attitude that Apple does that has prevented me from ever buying their products. Remember antennagate? "Not our fault. You are holding it wrong." I take it back, not chumps. Ass holes is more like it.
Universities often hold many patents. They do not often build products themselves from them. So by this narrow definition, you may see some relevance in calling them a patent troll.
However, Universities will normally license out those patents, even offering help in their implementation (for a fee). They will advertise them, not horde them looking for a payout later, and (AFAIK, I have never had to negotiate with them) their fees are generally much more reasonable. They do not just buy up patents from other people, but develop them in house.
In addition, the universities use the money from licensing to support their core businesses of teaching and R&D. The money is basically poured back in to developing ideas and minds.
Universities are (generally) not patent trolls. They are, in this area, R&D facilities pushing forward human knowledge.
That is bollox.....
when I signed up to do a course at Uni.... they requested I sign a paper saying anything I developed belong to them, including any research I might perform.
The fact that I paid them to enter Uni. did not seem to enter into it.
I told them they could go screw themselves.
they requested I sign a paper saying anything I developed belong to them, including any research I might perform.
This is fairly standard. At a uni, they are providing a lot of facilities for you. When I went, they allowed use for your own projects too.
If you have a good idea, it is often possible for you to talk to them and negotiate an exception, although they may expect something in return. You are paying them to teach you, not to help you found a new company, provide R&D equipment and resources etc.
The uni's boffins invented a table-like structure for the processor to keep track of data dependencies...
"invented" ?
Used a table to keep track...
Apple had countered that the patent was invalid and had suggested WARF was a patent troll.
Takes one...
...although I'm genuinely inclined to think they're right on this occasion. Should be too, considering they're one of the world's foremost authorities on patent trolling.
I think you'll all have to pay royalties to Apple because they invented the 'patent troll'. And using the phrase 'patent troll' will cost you, each and every time you read the phrase 'patent troll'.
And just be careful claiming anyone, apart from Apple, are 'patent trolls', cos that'll cost extra.
UW-Madison has long had one of the premier computer science departments in the US, right up there with some of the more well known names like MIT and several of the top California universities. I'd hardly call that a patent troll! I'd also expect them to have a considerable patent portfolio.
The patent system was supposed to encourage innovation, not to stop it dead. From all I have seen so far, the patent system seems to be more geared up to let people patent the obvious and then enrich lawyers fighting it out and on the odd occasion getting said patent invalidated. Meanwhile, the patent office has had its pound of flesh (for NOT doing its job), and the lawyers on either side had your limbs in the process.
Can you sue the patent office? It seems that the problem starts there, really.
The obviousness of what was patented can also go towards reducing the damage award. The idea was sufficiently obvious that the patent shouldn't be worth anywhere near $900 million.
Much of the reasoning behind the decision can be gleaned from the article's observation that "A jury, sitting in a US district court in Madison, Wisconsin, found Apple guilty ..."
"The idea was sufficiently obvious that ..."
Was it? I haven't read the patent in detail, but Intel sold *lots* of OoO chips before the patent was published and had both the incentive and the resources to make them as good as they could. However, *Intel* apparently settled out of court in a similar case a few years back. Perhaps this patent is for something clever. (Let's face it, cramming a worthwhile algorithm into a few transistors with next to no latency in decision making is going to involve *some* cleverness.)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
