Apple Says Qualcomm Has Overcharged Billions of Dollars By 'Double-Dipping' on iPhone's Innovation

[SarcasticJoe]

I'm actually glad to see someone finally take Qualcomm to court over their licensing practices...

Seriously, Qualcomm has been an absolute blight on the mobile industry for years and years. They own so many of the patents the CDMA standard is based on that if you want to make a CDMA device, they pretty much have you by the proverbial balls and they do a very good job at abusing this fact for all it's worth. Not only do you have to pay for the actual parts you put into your devices, you also have to pay them for the privilege to start designing a product with their parts and a percentage of the total sale price of each device you've sold originally set up when most phones were relatively inexpensive dumbphones and not smarphones costing as much as a decent home computer.

Apple is not the first company that's had enough of this over the years. Back when Nokia was still the #1 phone manufacturer by a long shot they also got tired of this and ended up getting sued by Qualcomm when they tried to get a better deal. That sadly ended in victory for Qualcomm, but I hope that this time Qualcomm doesn't get their way.

 

[techwhiz]

I'm actually glad to see someone finally take Qualcomm to court over their licensing practices...

Seriously, Qualcomm has been an absolute blight on the mobile industry for years and years. They own so many of the patents the CDMA standard is based on that if you want to make a CDMA device, they pretty much have you by the proverbial balls and they do a very good job at abusing this fact for all it's worth. Not only do you have to pay for the actual parts you put into your devices, you also have to pay them for the privilege to start designing a product with their parts and a percentage of the total sale price of each device you've sold originally set up when most phones were relatively inexpensive dumbphones and not smarphones costing as much as a decent home computer.

Apple is not the first company that's had enough of this over the years. Back when Nokia was still the #1 phone manufacturer by a long shot they also got tired of this and ended up getting sued by Qualcomm when they tried to get a better deal. That sadly ended in victory for Qualcomm, but I hope that this time Qualcomm doesn't get their way.

Qualcomm spent the money on R&D and created unique algorithms that make *ALL* cell phone communication possible.
I like Apple and hold shares in Apple, but they are wrong. They cry brand dilution when someone makes a phone that kind of looks like an iPhone, but claims that Qualcomm is abusive because they charge licensing for their innovation.
Explain that to me again?

Companies charge per chip royalties all the time. They charge an upfront cost of the chip then royalties on the products.
It's a common method to monetize technology. You charge based on the cost of the end product.
It allows even low cost products to enter the market.

Apple doesn't like the terms it agreed to because now they sell phones for more and they want to pay less.
They don't like the way Qualcomm licenses their technology, but they don't even license the iPhone.
Even if they move to Intel Modems, they will still need to pay Qualcomm.
Intel can't build a Modem without paying Qualcomm.
Qualcomm holds quite a few essential patents for CDMA and GSM.
https://www.greyb.com/starpatents-in-qualcomm-patent-portfolio/

Call it what you will, but without Veterbi and his algorithms, there would be no iPhone or anyother cell phone we see today. Qualcomm make the iPhone possible.

Doesn't Apple charge developers a percentage of App prices in the store?
They charge an up front fee for access to development kit and an ongoing subscription that you renew every year. They then charge a percentage of your App price. They also further constrain the market because even if you own the phone, you must use the App store. So they control the supply chain from end to end.

Explain to me again, how this is different?

 

[kdarling]

It's up for debate whether inertial scrolling is "transcendental" -- you can have functional scrolling without the "rubber-band" effect that was a key innovation of Apple's particular implementation of scrolling, which makes the scrolling feel more immediate to a user using their finger.

Just a note that Apple didn't invent inertial scrolling. It predates the iPhone by at least 15 years.

Similarly, with Multi-touch, it is up for debate whether it is "transcendental" ("natural", and no-one can possibly think of any other way) to, for example, pinch or separate two fingers to zoom in and out (or whatever Apple or its acquisitions such as FingerWorks may have come up with that hadn't been done before).

Again, multi-touch long predates Apple's usage. Fingerworks themselves acknowledge that they were based on many years of R&D before them.

The things Apple sued Samsung for are very specific.

Actually, a big problem was that some major ones were NOT very specific. Many of the design patents were vague, and the utility patents were not only seen by many old timers as non-patentable (and some were indeed later invalidated), but could also be interpreted in different ways (which is why it took more than one trial for some of them).

Until Intel produced a half decent Modem there was really no alternative but to use QC devices.

I'm pretty sure that there's not only Intel, but also MediaTek and Samsung modems available. In fact, Qualcomm chips sell less than those others in China.

There were even more alternatives until chipmakers stopped producing standalone modem chips, and began making integrated CPU + modem systems. Such all-in-one chipsets are much more in demand by phone makers.

The problem is that Apple won't use those, as they want to use their own CPU design instead.

Now there is an alternative and the barrel that QC had Apple (and all the other Device makers) over is suddenly an awful lot smaller.

True for Apple, but as noted just above, the majority of phone makers want integrated solutions to save time, money, space and power.

It's a good bet that Apple will want to join everyone else and fold Intel modems into their own SoCs.

Seriously, Qualcomm has been an absolute blight on the mobile industry for years and years. They own so many of the patents the CDMA standard is based on that if you want to make a CDMA device, they pretty much have you by the proverbial balls and they do a very good job at abusing this fact for all it's worth.

Mind you, I'm not disagreeing that QCOM charges as much as they can. But it's ironic that when Apple charges all it can, that's okay with many people. God forbid any other company try to do the same, though.

As for blight, that kind of ignores the fact that every phone on the planet uses QCOM CDMA technology for 3G. I'd say that gave a huge boost to smartphones at a critical time.

Not only do you have to pay for the actual parts you put into your devices, you also have to pay them for the privilege to start designing a product with their parts and a percentage of the total sale price of each device ...

Charging separately for silicon and software makes perfect sense. Chips constantly get cheaper to make, while the value of the IP needed to run on them stays constant.

 

[Carnegie]

Yet Apple knows full well that charging a percentage of product price has been standard cellular FRAND licensing for over two decades.



Apple itself is no stranger to the idea of charging a percentage of profits gained by using one's technology.

Is it also unfair that Apple charges higher priced apps a higher royalty, even though the same storage and download costs exist as for free apps?

Is it also unfair that Apple charges banks a percentage of each Apple Pay purchase, even though the chips and software being used have already been bought by a user at a high profit to Apple?



Nope, that never happens. Apple has no license from Qualcomm, and thus does not directly pay royalties themselves.

Qualcomm instead has a license agreement with Foxconn and the other iPhone assembly companies. The licenses cover many phone brands and predate Apple's iPhone. Those assembly companies pay a royalty for building devices which include 2G/3G/4G tech invented by Qualcomm.

And it's the same percentage whether it's an Apple or Moto phone, or whether it uses an Intel or Qualcomm or MediaTek modem. That's the fair and non-discriminatory parts of FRAND.

What Apple means by their disingenuous use of "double dip", is simply that the royalty Foxconn pays (and passes on to Apple) is higher for higher value iPhones. But see next response, because it's not based on anywhere near as much as Apple's retail or even wholesale prices. If anything, Apple has been getting by like a bandit for a decade by not paying on the full price.



Qualcomm's rates are well known, as hundreds of companies currently have licenses, all using the same basic terms, which is what FRAND requires. They charge ~3.25% of the device price.

However, in this case that means the price that Foxconn charges Apple for a boxed iPhone ready for Apple to make hundreds of dollars in profit from (something they could not do without using Qualcomm patents).

Analysts estimate that Foxconn charges Apple about $245 per phone, so the royalty would be about $8. And that's for thousands of essential patents without which the iPhone would only be an iPod.

Interestingly, Apple wanted that much per device from Samsung for just three Apple patents that ended up either being worked around or no longer used, like slide to unlock. In other words, non essential patents.

Apple is simply once again making its usual claim that its patents are worth tons, while everyone else's are worth mere pennies. It's just business, in the pursuit of higher profits. It's not like Apple would lower the price by $5 if they got a better royalty rate.

We've discussed the royalty base issue a number of times, so I won't go much into that here. But, again, the Federal Circuit has been clear on that issue and Qualcomm can't require that a would-be licensee agree to pay royalties (for FRAND-obliged patents) based on the entire market value of products. If a would-be licensee wants to do that, that's fine. But Qualcomm can't refuse to grant licenses because they won't agree to do that.

We've also discussed the per phone size of the royalties that Apple pays, so I won't get much into that here either. But the $245 per phone number, if it's meant to be an average, isn't accurate. We can say that with considerable confidence given what Apple reports.

Those things said, I'd like to respond to this:

Nope, that never happens. Apple has no license from Qualcomm, and thus does not directly pay royalties themselves.

Qualcomm instead has a license agreement with Foxconn and the other iPhone assembly companies. The licenses cover many phone brands and predate Apple's iPhone. Those assembly companies pay a royalty for building devices which include 2G/3G/4G tech invented by Qualcomm.

And it's the same percentage whether it's an Apple or Moto phone, or whether it uses an Intel or Qualcomm or MediaTek modem. That's the fair and non-discriminatory parts of FRAND.

What Apple means by their disingenuous use of "double dip", is simply that the royalty Foxconn pays (and passes on to Apple) is higher for higher value iPhones. But see next response, because it's not based on anywhere near as much as Apple's retail or even wholesale prices. If anything, Apple has been getting by like a bandit for a decade by not paying on the full price.

That is not what Apple means when it refers to Qualcomm double-dipping. We may disagree with Apple when it comes to the propriety of what it is (and others are) accusing Qualcomm of - whether it, e.g., is contrary to general patent law - but we should at least be clear about what those accusations are.

The double dipping that Apple is referring to is Qualcomm requiring Apple (through its contract manufacturers, of course) to pay licensing fees in addition to buying chipsets (which incorporate the technology which those licensing fees are supposedly for) from Qualcomm. The first sale doctrine means that, in buying those chipsets from Qualcomm, Qualcomm's patent rights are exhausted and it has no right to collect further licensing fees based on the technology within them. Apple alleges, as do others, that Qualcomm requires it (or its contract manufacturers, of course) to agree to (improper) separate licensing terms or else Qualcomm won't sell chipsets to Apple. Qualcomm is then paid for the chipsets and paid licensing fees. That's the core of the double-dipping accusation.

The higher royalties being paid because of the cost of iPhones being higher is a different issue, one of several other issues.

Connected with the double-dipping that I just described, is the issue of Apple having to pay licensing fees to Qualcomm even if it buys chipsets from Intel. Part of the problem is that Qualcomm - according to numerous parties, and as suggested by its own SEC filings - refuses to grant exhaustive licenses to chipset competitors such as Intel and Samsung. Those competitors have tried to get licenses from Qualcomm, and Qualcomm is supposed to grant them - that's part of its FRAND commitments. The way it should work is, those competitors would get licenses from Qualcomm (on FRAND terms) and then sell their chipsets (incorporating Qualcomm's SEPs) to the likes of Apple. Apple would then not need to pay Qualcomm any licensing fees for devices which use those chipsets purchased from Qualcomm's competitors. As it has been, Qualcomm has been requiring Apple (again, through its contract manufacturers) to pay it licensing fees even if Apple buys chipsets from, e.g., Intel. There are a number of reasons why Qualcomm won't license Intel and others, but one is that it wouldn't have the leverage to force them to agree to the high royalties it wants. (I'll go into why that's the case if anyone is interested.)

Further, you have the issue of Qualcomm not only making Apple pay royalties when it buys chipsets from Intel (which it shouldn't have to do as Intel should be able to get its own licenses) but effectively making Apple pay more in royalties if it buys chipsets from Intel rather than from Qualcomm. The accusation is that Qualcomm demands a higher than reasonable royalty, but then agrees to refund part of it if Apple buys chipsets from Qualcomm. So if Apple buys chipsets from Qualcomm's competitors, it has to pay more for the licensing rights. That aspect of Qualcomm's practices is likely also not FRAND-compliant. And it has the effect of discouraging companies like Apple from giving chipset business to Qualcomm's competitors; it artificially raises the effective price of Qualcomm's competitors' chipsets. There are multiple levels of wrong-doing involved even in just this particular aspect of the allegations made against Qualcomm.

At any rate, the main point I wanted to make is that the double-dipping that Apple accuses Qualcomm of is not what you are saying it is. Apple's filings make that clear, they spell out what Apple is alleging. Perhaps some will think the double-dipping in question is appropriate. If so, fair enough. But let's at least understand what the issue is. We should also understand that the Supreme Court has now made it clear that a patent holder is not, under patent law, entitled to collect royalties on patented items which it has already chosen to sell and it isn't, under patent law, allowed to place conditions on what may be done with them after it has sold them. That is to say, once a patent holder sells a patented item, further use of that item - to include reselling it - is not infringement, regardless of what conditions the patent holder may have placed on its original sale. If Apple (or, e.g., Foxconn) buys chipsets from Qualcomm, it isn't infringing Qualcomm's patents (which are incorporated in those chipsets) by putting them in iPhones and then selling those iPhones.

 

[rjohnstone]

Connected with the double-dipping that I just described, is the issue of Apple having to pay licensing fees to Qualcomm even if it buys chipsets from Intel. Part of the problem is that Qualcomm - according to numerous parties, and as suggested by its own SEC filings - refuses to grant exhaustive licenses to chipset competitors such as Intel and Samsung. Those competitors have tried to get licenses from Qualcomm, and Qualcomm is supposed to grant them - that's part of its FRAND commitments. The way it should work is, those competitors would get licenses from Qualcomm (on FRAND terms) and then sell their chipsets (incorporating Qualcomm's SEPs) to the likes of Apple. Apple would then not need to pay Qualcomm any licensing fees for devices which use those chipsets purchased from Qualcomm's competitors. As it has been, Qualcomm has been requiring Apple (again, through its contract manufacturers) to pay it licensing fees even if Apple buys chipsets from, e.g., Intel. There are a number of reasons why Qualcomm won't license Intel and others, but one is that it wouldn't have the leverage to force them to agree to the high royalties it wants. (I'll go into why that's the case if anyone is interested.)

Further, you have the issue of Qualcomm not only making Apple pay royalties when it buys chipsets from Intel (which it shouldn't have to do as Intel should be able to get its own licenses) but effectively making Apple pay more in royalties if it buys chipsets from Intel rather than from Qualcomm. The accusation is that Qualcomm demands a higher than reasonable royalty, but then agrees to refund part of it if Apple buys chipsets from Qualcomm. So if Apple buys chipsets from Qualcomm's competitors, it has to pay more for the licensing rights. That aspect of Qualcomm's practices is likely also not FRAND-compliant. And it has the effect of discouraging companies like Apple from giving chipset business to Qualcomm's competitors; it artificially raises the effective price of Qualcomm's competitors' chipsets. There are multiple levels of wrong-doing involved even in just this particular aspect of the allegations made against Qualcomm.

At any rate, the main point I wanted to make is that the double-dipping that Apple accuses Qualcomm of is not what you are saying it is. Apple's filings make that clear, they spell out what Apple is alleging. Perhaps some will think the double-dipping in question is appropriate. If so, fair enough. But let's at least understand what the issue is. We should also understand that the Supreme Court has now made it clear that a patent holder is not, under patent law, entitled to collect royalties on patented items which it has already chosen to sell and it isn't, under patent law, allowed to place conditions on what may be done with them after it has sold them. That is to say, once a patent holder sells a patented item, further use of that item - to include reselling it - is not infringement, regardless of what conditions the patent holder may have placed on its original sale. If Apple (or, e.g., Foxconn) buys chipsets from Qualcomm, it isn't infringing Qualcomm's patents (which are incorporated in those chipsets) by putting them in iPhones and then selling those iPhones.

Here is what I believe you are missing in the equation.
Qualcomm licenses are not blanket. There are manufacturing licenses and there are software licenses.

Foxconn has a license to manufacture QCOMM chips that are capable of performing patented functions.
Those functions are disabled by default.
To enable those functions, Foxconn requires the buyer of the chip to obtain a separate license from the patent holder (Qualcomm) to enable those functions.
This is pretty common practice on multifunction chipsets.

Foxconn would be out of their mind to obtain a license that covers both manufacturing the chip AND all of its capabilities. A lot of OEMs buy chips and do not want all of the features enabled as they simply do not need all of them.
The claim of patent exhaustion does not apply in this case.

In the past, Samsung for example, shipped phones that had a Qualcomm modem capable of using CDMA tech, but the tech was disabled on the devices that didn't need it. Thus no license was required for that feature.

 

[kdarling]

We've discussed the royalty base issue a number of times, so I won't go much into that here. But, again, the Federal Circuit has been clear on that issue and Qualcomm can't require that a would-be licensee agree to pay royalties (for FRAND-obliged patents) based on the entire market value of products. If a would-be licensee wants to do that, that's fine. But Qualcomm can't refuse to grant licenses because they won't agree to do that.

I think you're confusing a ruling about determining the base royalty for an infringed patent, with what is indeed a voluntary contract between companies.

As for the value of QCOM's IP, even the Chinese government's trade ministry... who can pretty much rule what they want... recently signed a deal where QCOM gets the same amount from Chinese phone makers as Foxconn pays on behalf of Apple. (Actually, they made it even MORE than what Apple pays, since it's based on retail price.)

We've also discussed the per phone size of the royalties that Apple pays, so I won't get much into that here either. But the $245 per phone number, if it's meant to be an average, isn't accurate. We can say that with considerable confidence given what Apple reports.

Okay, that's a base phone. If you upped the memory hugely, it still wouldn't add very much to the royalty, since as we all know too well, the extra memory cost that Apple would pay more royalty on, is a pittance compared to what Apple turns around and charges the end customer.

The double dipping that Apple is referring to is Qualcomm requiring Apple (through its contract manufacturers, of course) to pay licensing fees in addition to buying chipsets (which incorporate the technology which those licensing fees are supposedly for) from Qualcomm.

You're right, but it still doesn't happen. If the modem chips embodied the licensed patents, then yes patent exhaustion applies. But they don't.

Modem chipsets do NOT REPEAT NOT incorporate the technology that the licensing fees cover. Broadband modems are basically dumb DSPs with some custom circuitry. Anyone is allowed to design and make and sell such chips and compete with Qualcomm on the chip price. And indeed, in China other modem makers outsell Qualcomm because of their lower price.

And that's okay, because the QCOM license is not for the chip. It's for the software that is required to run on the chip to implement 2G/CDMA/4G/5G IP owned by QCOM.

The first sale doctrine means that, in buying those chipsets from Qualcomm, Qualcomm's patent rights are exhausted and it has no right to collect further licensing fees based on the technology within them.

See above. QCOM does NOT collect phonemaker licensing fees based on the technology in the chips themselves. That's a common layman confusion that Apple seeks to exploit.

I would also add that the new FTC Chief Commissioner has said that she saw no evidence that Qualcomm double dipped on anything.

Connected with the double-dipping that I just described, is the issue of Apple having to pay licensing fees to Qualcomm even if it buys chipsets from Intel. Part of the problem is that Qualcomm - according to numerous parties, and as suggested by its own SEC filings - refuses to grant exhaustive licenses to chipset competitors such as Intel and Samsung.

That is correct. QCOM only sells licenses to the device maker. That's who would ultimately pay the fees in ANY case. But this way, the chip makers don't have to also act as a fee collector.

It doesn't help that in the recent past, many chipmakers lied about how many chips they sold to phone makers in China, and thus QCOM lost out out tons in royalties. This is part of the recent China government decision as well, to help QCOM get paid.

The accusation is that Qualcomm demands a higher than reasonable royalty, but then agrees to refund part of it if Apple buys chipsets from Qualcomm.

That's more Apple handwaving, claiming that agreed upon kickbacks (which Apple happily collected for almost a decade) are suddenly horrible. There's also two sides presented here, and I think in trial which one is true will come out.

1. Apple claims that Qualcomm offered to refund (kickback) fees if Apple didn't use WiMAX in place.
2. Qualcomm claims that Apple threatened to use WiMAX instead of CDMA if QCOM didn't give them money.

 

[Carnegie]

Here is what I believe you are missing in the equation.
Qualcomm licenses are not blanket. There are manufacturing licenses and there are software licenses.

Foxconn has a license to manufacture QCOMM chips that are capable of performing patented functions.
Those functions are disabled by default.
To enable those functions, Foxconn requires the buyer of the chip to obtain a separate license from the patent holder (Qualcomm) to enable those functions.
This is pretty common practice on multifunction chipsets.

Foxconn would be out of their mind to obtain a license that covers both manufacturing the chip AND all of its capabilities. A lot of OEMs buy chips and do not want all of the features enabled as they simply do not need all of them.
The claim of patent exhaustion does not apply in this case.

In the past, Samsung for example, shipped phones that had a Qualcomm modem capable of using CDMA tech, but the tech was disabled on the devices that didn't need it. Thus no license was required for that feature.

Based on the portion of my post which you quoted and the arrangement which you are describing, I suspect that you didn't mean to refer to Foxconn but rather to Intel or another competitor to Qualcomm in the broadband processor market? If that's not correct, let me know and I'll try to respond with regard to Foxconn's situation.

Anyway, what you are describing is not something that I'm missing. Qualcomm - according to Apple and Intel and various other parties - refuses to grant complete exhaustive licenses to the likes of Intel. It doesn't want to do that for a number of reasons, to include that the likes of Intel would be in a better position to negotiate FRAND terms with Qualcomm. In other words, Qualcomm wouldn't be able to leverage them into agreeing to (what Apple and others consider) unreasonable licensing terms. The model that you describe is how Qualcomm prefers to arrange the situation (it has changed the way it refers to such situations over the years, but it now refers to them as "arrangements" in its SEC filings). And some would-be licensees might even prefer that kind of arrangement.

But, regardless of how Qualcomm wants to arrange the situation, it has an obligation to offer exhaustive licenses to would-be competitors. That is to say, it has an obligation to grant them licenses such that they can produce various (standards-incorporating) processors and then sell them to other parties (e.g. Apple) which could then use them in their products without having to pay Qualcomm additional licensing fees. Patent exhaustion would, in that case, apply. The point is that Qualcomm refuses to grant such license, even when would-be competitors want them.

 

[rjohnstone]

Based on the portion of my post which you quoted and the arrangement which you are describing, I suspect that you didn't mean to refer to Foxconn but rather to Intel or another competitor to Qualcomm in the broadband processor market? If that's not correct, let me know and I'll try to respond with regard to Foxconn's situation.

Anyway, what you are describing is not something that I'm missing. Qualcomm - according to Apple and Intel and various other parties - refuses to grant complete exhaustive licenses to the likes of Intel. It doesn't want to do that for a number of reasons, to include that the likes of Intel would be in a better position to negotiate FRAND terms with Qualcomm. In other words, Qualcomm wouldn't be able to leverage them into agreeing to (what Apple and others consider) unreasonable licensing terms. The model that you describe is how Qualcomm prefers to arrange the situation (it has changed the way it refers to such situations over the years, but it now refers to them as "arrangements" in its SEC filings). And some would-be licensees might even prefer that kind of arrangement.

But, regardless of how Qualcomm wants to arrange the situation, it has an obligation to offer exhaustive licenses to would-be competitors. That is to say, it has an obligation to grant them licenses such that they can produce various (standards-incorporating) processors and then sell them to other parties (e.g. Apple) which could then use them in their products without having to pay Qualcomm additional licensing fees. Patent exhaustion would, in that case, apply. The point is that Qualcomm refuses to grant such license, even when would-be competitors want them.

Correct... I was using Foxconn as an example, but it can apply to any chip maker.

As kdarling described, Qualcomm doesn't license the software portion to the chip maker. They never have.

 

[Carnegie]

I think you're confusing a ruling about determining the base royalty for an infringed patent, with what is indeed a voluntary contract between companies.

As for the value of QCOM's IP, even the Chinese government's trade ministry... who can pretty much rule what they want... recently signed a deal where QCOM gets the same amount from Chinese phone makers as Foxconn pays on behalf of Apple. (Actually, they made it even MORE than what Apple pays, since it's based on retail price.)



Okay, that's a base phone. If you upped the memory hugely, it still wouldn't add very much to the royalty, since as we all know too well, the extra memory cost that Apple would pay more royalty on, is a pittance compared to what Apple turns around and charges the end customer.



Still doesn't happen. That's Apple handwaving to the non-technical crowd.

Modem chipsets do NOT REPEAT NOT incorporate the technology that the licensing fees cover. Broadband modems are basically DSPs with some custom circuitry. Anyone is allowed to design and make and sell such chips and compete with Qualcomm on the chip price. (Intel bought a design that someone else derived from a Qualcomm licensed design, but again, that's just dumb silicon. It's like buying an enhanced ARM design.)

The QCOM license is not for the chip. It's for the software that is required to implement 2G/CDMA/4G/5G patents owned by QCOM.



See above. QCOM does NOT collect phonemaker licensing fees based on the technology in the chips themselves. That's a common layman confusion that Apple seeks to exploit.

I would also add that the new FTC Chief Commissioner has said that she saw no evidence that Qualcomm double dipped on anything.



That is correct. QCOM only sells license to the end user. That's who would pay the fees in ANY case. But this way, the chip makers don't have to also act as a fee collector.

It doesn't help that in the recent past, many chipmakers lied about how many chips they sold to phone makers in China, and thus QCOM lost out out tons in royalties. This is part of the recent China government decision as well, to help QCOM get paid.



That's more Apple handwaving, claiming that agreed upon kickbacks (which Apple happily collected for almost a decade) are suddenly horrible. There's also two sides presented here, and I think in trial which one is correct will come out. Because one of them isn't really believable.

1. Apple claims that Qualcomm offered to refund (kickback) fees if Apple didn't use WiMAX.
2. Qualcomm claims that Apple threatened to use WiMAX instead of CDMA if QCOM didn't give them money.

I'm not confusing the import of (or mistaking the relevance) the decisions which I've referred to. I've read them and am familiar with the issues they involved. A number of them related to FRAND-obliged patents. Further, they establish what is reasonable (under U.S. law) when it comes to compensation for ongoing use of patents and thus what's required (under U.S. law) when it comes to FRAND terms.

As I've indicated many times, when it comes to voluntary contracts between companies they are free to agree to royalties based on the value of end products. That's not what's at issue here. Here we're talking about terms which aren't mutually agreed to; we're talking about a FRAND-obligated patent holder seeking to impose terms which another party doesn't wish to agree to. Ultimately, if the impasse persists, a court would have to determine what appropriate royalties would be. And in so doing, under existing principles of U.S. law, they would not allow them to be based on the entire value of end products unless the patents in question could be demonstrated to account for the entire value of those products.

As for the dissenting FTC vote, Ms. Ohlhausen didn't specifically address the issue of double-dipping in her dissent - i.e., she didn't assert that there was no evidence that Qualcomm double dipped. She thought that the majority's legal theory was flawed (a position which I think has some merit). She also asserted that the complaint didn't even allege above-FRAND royalties, an assertion which was disingenuous. The complaint might not have used that exact phrasing, but substantively it did make that allegation. (Whether the allegation is correct is another matter, but Ms. Ohlhausen sought to bypass that consideration by pretending that the complaint didn't even make the allegation.) She also did a bit of handwaving when it came to whether it was alleged that Qualcomm effectively charged more for royalties when an OEM bought chips from someone other than Qualcomm. I've read hundreds, maybe thousands, of dissents. This one - while there were legitimate (dissenting) arguments to be made - was pretty unpersuasive.

More generally to the points you are making in response: As I indicated, we might disagree with Apple and various others when it comes to the allegations they've made. We may suspect that those allegations are false or we may believe that they don't refer to improper behavior. But my original point in responding was to correctly identify what those double-dipping allegations related to.
[doublepost=1498073410][/doublepost]

Correct... I was using Foxconn as an example, but it can apply to any chip maker.

As kdarling described, Qualcomm doesn't license the software portion to the chip maker. They never have.

I understand that. It has for many years referred to that arrangement in its SEC filings. The assertion (by Apple and Intel and others) is that it isn't allowed to do things that way - that it is required, if, e.g., Intel wants it to, to grant Intel an exhaustive license which would allow Intel to sell processors to Apple with all of Qualcomm's relevant patent rights exhausted. We can disagree with what's required, and on the veracity of the allegations made against Qualcomm, but that is one of the allegations.

Qualcomm of course prefers not to grant exhaustive licenses to would-be competitors.

 

[mi7chy]

Apple has a history of squeezing their suppliers to the point of bankruptcy. Apple went with Intel radio to squeeze more profit but it backfired since Intel radio reception and throughput are crappier. Apple now want to go back to Qualcomm but want Qualcomm to sell at a loss. From a national security point of view, Qualcomm's technology is much more valuable while Apple is just a front for Made in China. For this reason Qualcomm will prevail.

 

[kdarling]

As I've indicated many times, when it comes to voluntary contracts between companies they are free to agree to royalties based on the value of end products. That's not what's at issue here. Here we're talking about terms which aren't mutually agreed to; we're talking about a FRAND-obligated patent holder seeking to impose terms which another party doesn't wish to agree to. Ultimately, if the impasse persists, a court would have to determine what appropriate royalties would be.

FRAND is a voluntary commitment on the part of the patent holder. They do not lose normal patent rights nor the ability to set what they think is a fair rate. If enough others disagreed, then the patents would not get incorporated into the standard.

However, Qualcomm's CDMA design was so clearly superior to other tech of the time, that everyone immediately took licenses and even paid Qualcomm while also giving cross licenses. This shows how valuable Qualcomm's work was. Indeed, the entire world ended up using their CDMA technology for 3G.

Another point to consider is that cellular FRAND rules are set by the standards organization, in this case ETSI. Those rules include a way for a member to complain about not licensing to the ETSI Director, who can then convene a General Assembly to decide on a solution. If past history is any indication, Apple ignored those rules entirely in favor of trying to use US Courts as a rate bargaining method.

Finally, the US government has also put out rules and timelines for FRAND negotiations and arbitration that companies are supposed to follow. Has Apple followed them?

As for the dissenting FTC vote, Ms. Ohlhausen didn't specifically address the issue of double-dipping in her dissent -

You're right, what she said was that there was no evidence that Qualcomm charged customers more for using other chips.

She also said that it was not up to just customers to determine what is a "fair" rate, something that Apple loves to do, and which other officials have commented on before.

For example, even when judges in the past have offered to set a fair rate, Apple has refused to abide unless the rate was to their liking. Basically, Apple wants to be treated special, and that's not fair to other licensees.

... The assertion (by Apple and Intel and others) is that it isn't allowed to do things that way - that it is required, if, e.g., Intel wants it to, to grant Intel an exhaustive license which would allow Intel to sell processors to Apple with all of Qualcomm's relevant patent rights exhausted.

Since this is about FRAND commitments as defined by ETSI, their rules apply. Now, ETSI patent licensing rules apply to allowing others to manufacture and use patented methods. Qualcomm does license such chip patents to chipmakers. Heck, Intel has such a license !!!!

But what Qualcomm does not freely license to chipmakers are their software copyrights. Qualcomm says this means those licenses are not subject to FRAND rules for patents, and that's true.

Nevertheless they would be subject to ETSI FRAND rules for software... IF Qualcomm had also offered their software to ETSI as part of the standard. Interestingly, I'm not sure they did so.

If not, then there's nothing a court can decide upon here, since what Apple is really complaining about, is copyrighted software that might not be under any FRAND commitment.

But I could be wrong. It'll be interesting to follow any trial.

 

[cmaier]

Yet Apple knows full well that charging a percentage of product price has been standard cellular FRAND licensing for over two decades.

"a] patentee may assess damages based on the entire market value of the accused product only where the patented feature creates the basis for customer demand or substantially creates the value of the component parts.'” Id., quoting Versata Software, Inc. v. SAP Am., Inc., 717 F.3d 1255, 1268 (Fed. Cir. 2013)

Of course, it's not so simple...

 

[kdarling]

"a] patentee may assess damages based on the entire market value of the accused product only where the patented feature creates the basis for customer demand or substantially creates the value of the component parts.'” Id., quoting Versata Software, Inc. v. SAP Am., Inc., 717 F.3d 1255, 1268 (Fed. Cir. 2013)

Sure, if assessing infringement damage from scratch, where there's no other guide to value.

However, in this case the court has available the same basic rates paid by hundreds of previous and current licensees over something like 35 years.

"Where an established royalty rate for the patented inventions is shown to exist, the rate will usually be adopted as the best measure of reasonable and entire compensation.” Tektronix vs. US, Fed Circuit 1977

 

[democracyrules]

They both make good points. I'm a bit tired of the constant lawsuits involving Apple, but hopefully this doesn't take years (à la Samsung) to wade through, and the courts can issue a ruling reasonably quickly.

EDIT: Having re-read the other MR article about Qualcomm's counter suit, I think I side more with Apple on this, instead of being neutral. Qualcomm's arguments don't really seem to be arguments, more or less just going on about how "iPhone wouldn't have been possible without us!" and claiming Apple didn't try to come to fair agreements, but there is no evidence provided (at least in the MR article).

Again, I guess we just need to let the courts do their job and listen to both parties before making a ruling.

Totally agree with the excellent comment above. Apple is on the right position.

 

[ravenstar]

Problem is you do have to use Apple's OS.

I agree with you though on all the other points, except at the end. As far as I last saw, there is absolutely no way to load an alternative OS onto an iPhone. Even if you wanted to write one.

we buy the hardware. We should have exclusive right to do whatever we want to it and install any software we want to it we own the right to. The manufacturers should warn that it's advanced to do so and could void warranties, But it is still my hardware and I should have the right to bust my device doing something stupid if I so shall please.

Are you really saying that because a product contains software it must be designed to allow you to put whatever software you wish on it? Your statement makes no sense at all.

 

[cmaier]

Sure, if assessing infringement damage from scratch, where there's no other guide to value.

However, in this case the court has available the same basic rates paid by hundreds of previous and current licensees over something like 35 years.

"Where an established royalty rate for the patented inventions is shown to exist, the rate will usually be adopted as the best measure of reasonable and entire compensation.” Tektronix vs. US, Fed Circuit 1977

My case is much newer ;-)

Like I said, it's complicated. Unfortunately i can't say what I really think.

 

[LordVic]

Are you really saying that because a product contains software it must be designed to allow you to put whatever software you wish on it? Your statement makes no sense at all.

to clarify:

If I buy a piece of hardware, I should have the right to completely remove all traces of any software on it, including bundled OS, and reinstall that OS, or any other OS of my choosing, should one exist.

 

[Michael Goff]

to clarify:

If I buy a piece of hardware, I should have the right to completely remove all traces of any software on it, including bundled OS, and reinstall that OS, or any other OS of my choosing, should one exist.

That’s unworkable and naive.

 

[LordVic]

That’s unworkable and naive.

I'd like to think more idealistic. And I will always put that into consideration when I purchase a device. Does it mean I'll replace the OS all the time? no, But I believe I, and everyone else who spends their money on these things, deserve. I'm not renting the hardware. I bought it. it's mine. The software I am using on license. I should be able to choose not to use that software if I decline the license terms, and use another piece of software instead that I do agree with the terms.

in modern phones, and it's not just Apple doing it anymore, if you decline the software terms of services for the device, that's it. The device is bricked until you do, or find someone who will buy it and agree. I do not agree with companies having the right to do this once we have purchased the hardware.

 

[Michael Goff]

I'd like to think more idealistic. And I will always put that into consideration when I purchase a device. Does it mean I'll replace the OS all the time? no, But I believe I, and everyone else who spends their money on these things, deserve. I'm not renting the hardware. I bought it. it's mine. The software I am using on license. I should be able to choose not to use that software if I decline the license terms, and use another piece of software instead that I do agree with the terms.

in modern phones, and it's not just Apple doing it anymore, if you decline the software terms of services for the device, that's it. The device is bricked until you do, or find someone who will buy it and agree. I do not agree with companies having the right to do this once we have purchased the hardware.

So you want to be able to change the OS in your car? And do you realize the amount of work required from everyone to make every OS work with every piece of hardware?

 

[LordVic]

So you want to be able to change the OS in your car? And do you realize the amount of work required from everyone to make every OS work with every piece of hardware?

I should be able to change the software in my car. Yes.

and I do not expect every software company to need to support every hardware ever. That's not the point. it's up to the buyer (myself) to be aware when I purchase what the free open market for software has come up with for that device. If there's nothing else, than so be it. that's just a factor in a purchasing decision.

But it annoys me when hardware manufacturers purposely block it. locking that hardware exclusively to them. it's basically extortion. "Use our software, or you can't use your hardware".

 

[Michael Goff]

I should be able to change the software in my car. Yes.

and I do not expect every software company to need to support every hardware ever. That's not the point. it's up to the buyer (myself) to be aware when I purchase what the free open market for software has come up with for that device. If there's nothing else, than so be it. that's just a factor in a purchasing decision.

But it annoys me when hardware manufacturers purposely block it. locking that hardware exclusively to them. it's basically extortion. "Use our software, or you can't use your hardware".

Extortion? Don’t you think you’re being hyperbolic? I don’t go into McDonalds expecting a whopper and I don’t go to Apple expecting Android.

I think you’re being unrealistic. Idealism is great and all, but do try to live in the real world.

 

[LordVic]

Extortion? Don’t you think you’re being hyperbolic? I don’t go into McDonalds expecting a whopper and I don’t go to Apple expecting Android.

I think you’re being unrealistic. Idealism is great and all, but do try to live in the real world.

of course i'm being hyperbolic. I still buy the stuff. As I said, it's a factor. I preffer it open. But it's not the only factor. It's not some ideolistic crusade, where I boycot everything and everyone who isn't true to my ideals.

I've also experienced in the past where I've bought a closed product like that and got burned horribly (Blackberry PLaybook, its ok, you can laugh). So it's not like I don't speak from experience. I still buy Apple stuff, because of other redeeming factors that usually make up for it.

 

[Michael Goff]

of course i'm being hyperbolic. I still buy the stuff. As I said, it's a factor. I preffer it open. But it's not the only factor. It's not some ideolistic crusade, where I boycot everything and everyone who isn't true to my ideals.

I've also experienced in the past where I've bought a closed product like that and got burned horribly (Blackberry PLaybook, its ok, you can laugh). So it's not like I don't speak from experience. I still buy Apple stuff, because of other redeeming factors that usually make up for it.

Why would I laugh about that? I tried to buy a Playbook back in the day and had to return it because Best Buy had sold me a 150$ empty box.

 

[kdarling]

So you want to be able to change the OS in your car? And do you realize the amount of work required from everyone to make every OS work with every piece of hardware?

I think you're overstating what he wishes to be able to do.

He's not asking hardware companies to support other OSes. He's simply asking that they don't deliberately block them.