Qualcomm must license modem tech to rivals like Intel, court rules

2»

Comments

  • Reply 21 of 39
    StrangeDaysStrangeDays Posts: 12,834member
    Gaby said:
    gatorguy said:
    noelos said:
    gatorguy said:
    chasm said:
    Would be interesting to know if Koh forced “QUALCOMM” to license those patents on FRAND terms or not.
    What are "FRAND terms"? Serious question to you.


    http://lmgtfy.com/?q=frand+patent+licensing


    Cute, but doesn't answer the question of what the OP means by FRAND terms. I already have a good idea what is allowed and what isn't. Since you chimed in and know how to use a search box do you personally think licensing based on a percentage of end product cost or revenues can meet the criteria for FRAND-compliant? That seems to be a big sticking point for many here. 
    Of course not! With all due respect the answer is in the acronym - fair reasonable and non discriminatory. To charge different prices based on whether or not a product is “premium” is by its very definition discriminatory! If you have a component or patent for sale/licensing, it’s ludicrous to then say yes we will sell it to you but can you tell me the price of your product please so we can decide how much to charge. 
    I mean can you imagine walking into a shop or supermarket  and being confronted by someone requesting your financials and annual salary so they could work out how much to charge you?!  😂
    Yup exactly. It’s ridiculous. 
    Gaby
  • Reply 22 of 39
    The value of the phone that the chip goes into should have no bearing on the licensing cost. That would be like charging more for a gallon of gas going into a Ferrari vs a Toyota. 

    Not sure if volume licensing discounts are permissible under FRAND. I believe they might be, as long as they’re consistent. 
    Gaby
  • Reply 23 of 39
    gatorguygatorguy Posts: 24,176member
    Gaby said:
    gatorguy said:
    noelos said:
    gatorguy said:
    chasm said:
    Would be interesting to know if Koh forced “QUALCOMM” to license those patents on FRAND terms or not.
    What are "FRAND terms"? Serious question to you.


    http://lmgtfy.com/?q=frand+patent+licensing


    Cute, but doesn't answer the question of what the OP means by FRAND terms. I already have a good idea what is allowed and what isn't. Since you chimed in and know how to use a search box do you personally think licensing based on a percentage of end product cost or revenues can meet the criteria for FRAND-compliant? That seems to be a big sticking point for many here. 
    Of course not!  ߘ⦬t;/div>
    ...Except that it is, and commonly so. Look no further than 3G, or 4G or 5G standards royalty rates. Do a bit of research on how Ericsson, or Nokia, or Huawei, or LG or a half dozen other companies involved in smartphone tech establish and collect royalties for their SEP-pledged IP. 

    So "Fair": Fair to both the owner of the intellectual property and the potential licensee? Only fair to one of them? 
    "Reasonable": Again, reasonable to who and who decides that? The standards group, ie ETSI or IEEE? The licensee?
    "Non-Discriminatory": Is it discriminatory when all licensees are bound to the same percentage of sales? 

    That's the issue with "FRAND". None of the terms are clearly defined, nor is the methodology for determining when the terms are met. No court has ruled that rates based on revenues are plainly non-FRAND AFAIK. 

    edited November 2018
  • Reply 24 of 39
    gatorguygatorguy Posts: 24,176member
    chasm said:
    Would be interesting to know if Koh forced “QUALCOMM” to license those patents on FRAND terms or not.
    So @chasm ; I had expected you might have some opinion of what "FRAND terms" means to you, but no matter.  The crux of the ruling if you read it is that Qualcomm will be required to license on "reasonable" (FRAND) terms and per the commitment they made to two US standards agencies. The exact rate to be considered "reasonable" was not stipulated by Judge Koh as far as I saw so likely treated the same as any other customer licensing their SEP package of patents. 
    edited November 2018
  • Reply 25 of 39
    carnegiecarnegie Posts: 1,077member
    gatorguy said:
    chasm said:
    Would be interesting to know if Koh forced “QUALCOMM” to license those patents on FRAND terms or not.
    So @chasm ; I had expected you might have some opinion of what "FRAND terms" means to you, but no matter.  The crux of the ruling if you read it is that Qualcomm will be required to license on "reasonable" (FRAND) terms and per the commitment they made to two US standards agencies. The exact rate to be considered "reasonable" was not stipulated by Judge Koh as far as I saw so likely treated the same as any other customer licensing their SEP package of patents. 
    Correct, Judge Koh didn't dictate specific terms on which Qualcomm has to license SEPs to its modem competitors. The only point of this motion was to get a finding that Qualcomm did indeed, in accordance with its contractual obligations (relating to SEPs and SSOs), have to license to its modem competitors. Qualcomm disputed that it did. The FTC wanted to tie up that particular string before trial so as to eliminate it as something which might add unnecessary complexity to the trial.

    It's something Qualcomm really should have stipulated to. But that would have amounted to admitting that it had been violating its contractual obligations all along. So it took a laughable position on this particular legal point, and Judge Koh had to waste time considering the matter. If Qualcomm were correct in its position, that could cause all kinds of problems for industry standards. It would also mean that other parties which hold SEPs could refuse to license to Qualcomm the IP which Qualcomm needs to make and sell modems. But Qualcomm knew  - at least, I sure hope it knew - that it would lose on this motion, so it could afford to take the position it did without fear of such consequences.
    chasm
  • Reply 26 of 39
    GabyGaby Posts: 190member
    gatorguy said:
    Gaby said:
    gatorguy said:
    noelos said:
    gatorguy said:
    chasm said:
    Would be interesting to know if Koh forced “QUALCOMM” to license those patents on FRAND terms or not.
    What are "FRAND terms"? Serious question to you.


    http://lmgtfy.com/?q=frand+patent+licensing


    Cute, but doesn't answer the question of what the OP means by FRAND terms. I already have a good idea what is allowed and what isn't. Since you chimed in and know how to use a search box do you personally think licensing based on a percentage of end product cost or revenues can meet the criteria for FRAND-compliant? That seems to be a big sticking point for many here. 
    Of course not!  ߘ⦬t;/div>
    ...Except that it is, and commonly so. Look no further than 3G, or 4G or 5G standards royalty rates. Do a bit of research on how Ericsson, or Nokia, or Huawei, or LG or a half dozen other companies involved in smartphone tech establish and collect royalties for their SEP-pledged IP. 

    So "Fair": Fair to both the owner of the intellectual property and the potential licensee? Only fair to one of them? 
    "Reasonable": Again, reasonable to who and who decides that? The standards group, ie ETSI or IEEE? The licensee?
    "Non-Discriminatory": Is it discriminatory when all licensees are bound to the same percentage of sales? 

    That's the issue with "FRAND". None of the terms are clearly defined, nor is the methodology for determining when the terms are met. No court has ruled that rates based on revenues are plainly non-FRAND AFAIK. 

    This is part of the problem with using such ambiguous language when creating these rules. It’s the same in many contracts and other legal writings. It’s about time that it’s reviewed and standardised in my opinion. It’s like ISPs’ “fair usage policies” or unlimited mobile data allowing “reasonable” limits. Very poorly defined, subjective, and entirely open to interpretation. Moreover it allows more unscrupulous judges to do as they please without recourse. 

    But if it is really the case that legally speaking they can pick and choose what to charge based on end price of the product or whatever arbitrary parameters they choose to apply, then it’s a very flawed and bent system indeed 
    edited November 2018
  • Reply 27 of 39
    chasmchasm Posts: 3,274member
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow, but the key element of any actual FRAND compliance would be to stop trying to charge a percentage of the entire unit (phone) rather than a price based on your percentage of said phone. For starters.

    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    edited November 2018
  • Reply 28 of 39
    gatorguygatorguy Posts: 24,176member
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
  • Reply 29 of 39
    carnegiecarnegie Posts: 1,077member
    gatorguy said:
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
    Apple won't necessarily have to pay Qualcomm royalties, based on the entire market value, when it doesn't use Qualcomm hardware (and even when it does, for that matter). That's part of the import of the ruling which is the subject of this thread.

    Apple might agree to pay royalties directly to Qualcomm on the entire market value, if the terms of its agreement with Qualcomm are otherwise favorable enough. But Apple has the a lot leverage now when it comes to whether it agrees to do so and on what terms it might agree to do so.

    One of the things which Qualcomm used to do, which violated its FRAND obligations, was refuse to license directly to its modem competitors. That allowed it to do some of the other improper things that it did. That's how Qualcomm's scheme worked, it employed a number of (illegal or contract violative) tactics which all worked together and which wouldn't have been effective on their own.

    If Qualcomm licensed to, e.g. Intel, directly, then Apple wouldn't need to pay Qualcomm for the IP in the modems which it buys from Intel. Qualcomm's IP rights would be exhausted. So it wouldn't have a way to charge Apple based on the entire market value (of end products). Whatever Intel pays wouldn't be based on the value of the devices its modems go into. At most, it would be based on the value of the modems (and even that wouldn't necessarily be the case, based on recent Federal Circuit decisions). This is why Qualcomm had previously refused to license to, e.g., Intel.

    When it comes to other Qualcomm IP which Apple might want to license - IP which isn't FRAND encumbered - Qualcomm's case for requiring royalties based on the entire market value has only gotten weaker. The Federal Circuit had already been clear that patent holders weren't entitled to royalties based on the entire market value unless certain conditions were met, and it's safe to say that those conditions weren't met with regard to, e.g., iPhones. But with its recent decision in Power Integrations v Fairchild Semiconductor International, the Federal Circuit arguably made it even more difficult for patent holders to establish that they were entitled to royalties based on the entire market value (citations omitted):

    As LaserDynamics, Versata, and VirnetX held, the entire market value rule is appropriate only when the patented feature is the sole driver of customer demand or substantially creates the value of the component parts. The burden of proof in this respect is on the patent holder. The question is whether the accused product, compared to other products in the same field, has features that would cause consumers to purchase the products beyond the patented feature, i.e., valuable features. Where the accused infringer presents evidence that its accused product has other valuable features beyond the patented feature, the patent holder must establish that these features do not cause consumers to purchase the product. A patentee may do this by showing that the patented feature “alone motivates customers to purchase [the infringing product]” in the first place. But when the product contains multiple valuable features, it is not enough to merely show that the patented feature is viewed as essential, that a product would not be commercially viable without the patented feature, or that consumers would not purchase the product without the patented feature. When the product contains other valuable features, the patentee must prove that those other features do not cause consumers to purchase the product. 
    edited November 2018
  • Reply 30 of 39
    gatorguygatorguy Posts: 24,176member
    carnegie said:
    gatorguy said:
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
    Apple won't necessarily have to pay Qualcomm royalties, based on the entire market value, when it doesn't use Qualcomm hardware (and even when it does, for that matter). That's part of the import of the ruling which is the subject of this thread.

    Apple might agree to pay royalties directly to Qualcomm on the entire market value, if the terms of its agreement with Qualcomm are otherwise favorable enough. But Apple has the a lot leverage now when it comes to whether it agrees to do so and on what terms it might agree to do so.

    One of the things which Qualcomm used to do, which violated its FRAND obligations, was refuse to license directly to its modem competitors. That allowed it to do some of the other improper things that it did. That's how Qualcomm's scheme worked, it employed a number of (illegal or contract violative) tactics which all worked together and which wouldn't have been effective on their own.

    If Qualcomm licensed to, e.g. Intel, directly, then Apple wouldn't need to pay Qualcomm for the IP in the modems which it buys from Intel. Qualcomm's IP rights would be exhausted. So it wouldn't have a way to charge Apple based on the entire market value (of end products). Whatever Intel pays wouldn't be based on the value of the devices its modems go into. At most, it would be based on the value of the modems (and even that wouldn't necessarily be the case, based on recent Federal Circuit decisions). This is why Qualcomm had previously refused to license to, e.g., Intel.

    When it comes to other Qualcomm IP which Apple might want to license - IP which isn't FRAND encumbered - Qualcomm's case for requiring royalties based on the entire market value has only gotten weaker. The Federal Circuit had already been clear that patent holders weren't entitled to royalties based on the entire market value unless certain conditions were met, and it's safe to say that those conditions weren't met with regard to, e.g., iPhones. But with its recent decision in Power Integrations v Fairchild Semiconductor International, the Federal Circuit arguably made it even more difficult for patent holders to establish that they were entitled to royalties based on the entire market value (citations omitted):

    As LaserDynamics, Versata, and VirnetX held, the entire market value rule is appropriate only when the patented feature is the sole driver of customer demand or substantially creates the value of the component parts. The burden of proof in this respect is on the patent holder. The question is whether the accused product, compared to other products in the same field, has features that would cause consumers to purchase the products beyond the patented feature, i.e., valuable features. Where the accused infringer presents evidence that its accused product has other valuable features beyond the patented feature, the patent holder must establish that these features do not cause consumers to purchase the product. A patentee may do this by showing that the patented feature “alone motivates customers to purchase [the infringing product]” in the first place. But when the product contains multiple valuable features, it is not enough to merely show that the patented feature is viewed as essential, that a product would not be commercially viable without the patented feature, or that consumers would not purchase the product without the patented feature. When the product contains other valuable features, the patentee must prove that those other features do not cause consumers to purchase the product. 
    You could well be correct a few years down the road, but with several lawsuits over Qualcomm licensing terms settled and others underway has any court yet said that royalties based on a total device cost is illegal or even un-FRAND-ly? 
  • Reply 31 of 39
    carnegiecarnegie Posts: 1,077member
    gatorguy said:
    carnegie said:
    gatorguy said:
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
    Apple won't necessarily have to pay Qualcomm royalties, based on the entire market value, when it doesn't use Qualcomm hardware (and even when it does, for that matter). That's part of the import of the ruling which is the subject of this thread.

    Apple might agree to pay royalties directly to Qualcomm on the entire market value, if the terms of its agreement with Qualcomm are otherwise favorable enough. But Apple has the a lot leverage now when it comes to whether it agrees to do so and on what terms it might agree to do so.

    One of the things which Qualcomm used to do, which violated its FRAND obligations, was refuse to license directly to its modem competitors. That allowed it to do some of the other improper things that it did. That's how Qualcomm's scheme worked, it employed a number of (illegal or contract violative) tactics which all worked together and which wouldn't have been effective on their own.

    If Qualcomm licensed to, e.g. Intel, directly, then Apple wouldn't need to pay Qualcomm for the IP in the modems which it buys from Intel. Qualcomm's IP rights would be exhausted. So it wouldn't have a way to charge Apple based on the entire market value (of end products). Whatever Intel pays wouldn't be based on the value of the devices its modems go into. At most, it would be based on the value of the modems (and even that wouldn't necessarily be the case, based on recent Federal Circuit decisions). This is why Qualcomm had previously refused to license to, e.g., Intel.

    When it comes to other Qualcomm IP which Apple might want to license - IP which isn't FRAND encumbered - Qualcomm's case for requiring royalties based on the entire market value has only gotten weaker. The Federal Circuit had already been clear that patent holders weren't entitled to royalties based on the entire market value unless certain conditions were met, and it's safe to say that those conditions weren't met with regard to, e.g., iPhones. But with its recent decision in Power Integrations v Fairchild Semiconductor International, the Federal Circuit arguably made it even more difficult for patent holders to establish that they were entitled to royalties based on the entire market value (citations omitted):

    As LaserDynamics, Versata, and VirnetX held, the entire market value rule is appropriate only when the patented feature is the sole driver of customer demand or substantially creates the value of the component parts. The burden of proof in this respect is on the patent holder. The question is whether the accused product, compared to other products in the same field, has features that would cause consumers to purchase the products beyond the patented feature, i.e., valuable features. Where the accused infringer presents evidence that its accused product has other valuable features beyond the patented feature, the patent holder must establish that these features do not cause consumers to purchase the product. A patentee may do this by showing that the patented feature “alone motivates customers to purchase [the infringing product]” in the first place. But when the product contains multiple valuable features, it is not enough to merely show that the patented feature is viewed as essential, that a product would not be commercially viable without the patented feature, or that consumers would not purchase the product without the patented feature. When the product contains other valuable features, the patentee must prove that those other features do not cause consumers to purchase the product. 
    You could well be correct a few years down the road, but with several lawsuits over Qualcomm licensing terms settled and others underway has any court yet said that royalties based on a total device cost is illegal or even un-FRAND-ly? 
    My comments were based on the state of the law as it is today, not as it might be in the future.

    As I've indicated before, royalties based on the entire market value of devices aren't illegal. Parties can freely agree to such terms. Apple could, e.g., decide that the rest of the terms of a comprehensive agreement with Qualcomm were favorable enough that it would be willing to pay royalties to Qualcomm based on the entire market value of its devices.

    But when it comes to what Qualcomm is entitled to - e.g., what it should be awarded by a court if it can't agree with Apple on what proper royalties should be - the Federal Circuit has been clear on this point. Patent holders are not entitled to royalties based on the entire market value unless they can demonstrate that their patents create all of the demand form the products in question, that other features of those products don't drive any of the demand for them. And when it comes to SEPs in particular, the Federal Circuit has also been clear in that the value that their holders' are entitled to is that intrinsic to the patented inventions - not that which is created by those inventions being included in the relevant standards.

    In deciding patent royalty cases, and setting out the rules which are to apply to damage calculations, the Federal Circuit is deciding what is reasonable in terms of what is owed to patent holders. FRAND terms, as you are aware, generally aren't specified in detail in the agreements which standard-setting organizations make with SEP holders. What represents and doesn't represent FRAND terms is determined by industry practices and case law. The latter, to the extent it exists on a particular point, is what controls.

    The Federal Circuit saying which methods are proper when it comes to calculating royalties owed to patent holders - in cases where parties don't agree - determines what is reasonable when it comes to SEPs bound by FRAND terms. Relevant here, even if the Federal Circuit hadn't applied its entire market value rule in cases that specifically dealt with FRAND-bound patents, that rule would still dictate what was proper in such cases. The general rule would still apply.

    If anything, it would be going in the other direction where an argument could have been made that the rule didn't apply. If the court had only ever ruled that it applied in cases involving FRAND-bound patents, then it might be argued whether it also applies in cases which didn't involve FRAND-bound patents. Indeed, that was one of the issues addressed in CSIRO v Cisco (Federal Circuit, 2015). Do the rules from Ericsson apply even if the patents in question don't carry FRAND obligations?

    But, to be clear: Ericsson v D-Link (2014) - a case which addressed a number of legal issues - involved patents that carried FRAND obligations. And in that case the Federal Circuit reiterated the entire market value rule which I've been referring to. It said, among other things (citations omitted):

    Thus, where the entire value of a machine as a marketable article is “properly and legally attributable to the patented feature,” the damages owed to the patentee may be calculated by reference to that value. Where it is not, however, courts must insist on a more realistic starting point for the royalty calculations by juries—often, the smallest salable unit and, at times, even less.
    The Federal Circuit has now reiterated this rule in a number of cases and contexts. And it follows from more general Supreme Court precedent that's more than a century old. That is the state of the law as it exists now. It could, of course, change in the future. 



  • Reply 32 of 39
    gatorguygatorguy Posts: 24,176member
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
    Apple won't necessarily have to pay Qualcomm royalties, based on the entire market value, when it doesn't use Qualcomm hardware (and even when it does, for that matter). That's part of the import of the ruling which is the subject of this thread.

    Apple might agree to pay royalties directly to Qualcomm on the entire market value, if the terms of its agreement with Qualcomm are otherwise favorable enough. But Apple has the a lot leverage now when it comes to whether it agrees to do so and on what terms it might agree to do so.

    One of the things which Qualcomm used to do, which violated its FRAND obligations, was refuse to license directly to its modem competitors. That allowed it to do some of the other improper things that it did. That's how Qualcomm's scheme worked, it employed a number of (illegal or contract violative) tactics which all worked together and which wouldn't have been effective on their own.

    If Qualcomm licensed to, e.g. Intel, directly, then Apple wouldn't need to pay Qualcomm for the IP in the modems which it buys from Intel. Qualcomm's IP rights would be exhausted. So it wouldn't have a way to charge Apple based on the entire market value (of end products). Whatever Intel pays wouldn't be based on the value of the devices its modems go into. At most, it would be based on the value of the modems (and even that wouldn't necessarily be the case, based on recent Federal Circuit decisions). This is why Qualcomm had previously refused to license to, e.g., Intel.

    When it comes to other Qualcomm IP which Apple might want to license - IP which isn't FRAND encumbered - Qualcomm's case for requiring royalties based on the entire market value has only gotten weaker. The Federal Circuit had already been clear that patent holders weren't entitled to royalties based on the entire market value unless certain conditions were met, and it's safe to say that those conditions weren't met with regard to, e.g., iPhones. But with its recent decision in Power Integrations v Fairchild Semiconductor International, the Federal Circuit arguably made it even more difficult for patent holders to establish that they were entitled to royalties based on the entire market value (citations omitted):

    As LaserDynamics, Versata, and VirnetX held, the entire market value rule is appropriate only when the patented feature is the sole driver of customer demand or substantially creates the value of the component parts. The burden of proof in this respect is on the patent holder. The question is whether the accused product, compared to other products in the same field, has features that would cause consumers to purchase the products beyond the patented feature, i.e., valuable features. Where the accused infringer presents evidence that its accused product has other valuable features beyond the patented feature, the patent holder must establish that these features do not cause consumers to purchase the product. A patentee may do this by showing that the patented feature “alone motivates customers to purchase [the infringing product]” in the first place. But when the product contains multiple valuable features, it is not enough to merely show that the patented feature is viewed as essential, that a product would not be commercially viable without the patented feature, or that consumers would not purchase the product without the patented feature. When the product contains other valuable features, the patentee must prove that those other features do not cause consumers to purchase the product. 
    You could well be correct a few years down the road, but with several lawsuits over Qualcomm licensing terms settled and others underway has any court yet said that royalties based on a total device cost is illegal or even un-FRAND-ly? 
    My comments were based on the state of the law as it is today, not as it might be in the future.

    As I've indicated before, royalties based on the entire market value of devices aren't illegal. Parties can freely agree to such terms. Apple could, e.g., decide that the rest of the terms of a comprehensive agreement with Qualcomm were favorable enough that it would be willing to pay royalties to Qualcomm based on the entire market value of its devices.

    But when it comes to what Qualcomm is entitled to - e.g., what it should be awarded by a court if it can't agree with Apple on what proper royalties should be - the Federal Circuit has been clear on this point. Patent holders are not entitled to royalties based on the entire market value unless they can demonstrate that their patents create all of the demand form the products in question, that other features of those products don't drive any of the demand for them. And when it comes to SEPs in particular, the Federal Circuit has also been clear in that the value that their holders' are entitled to is that intrinsic to the patented inventions - not that which is created by those inventions being included in the relevant standards.

    In deciding patent royalty cases, and setting out the rules which are to apply to damage calculations, the Federal Circuit is deciding what is reasonable in terms of what is owed to patent holders. FRAND terms, as you are aware, generally aren't specified in detail in the agreements which standard-setting organizations make with SEP holders. What represents and doesn't represent FRAND terms is determined by industry practices and case law. The latter, to the extent it exists on a particular point, is what controls.

    The Federal Circuit saying which methods are proper when it comes to calculating royalties owed to patent holders - in cases where parties don't agree - determines what is reasonable when it comes to SEPs bound by FRAND terms. Relevant here, even if the Federal Circuit hadn't applied its entire market value rule in cases that specifically dealt with FRAND-bound patents, that rule would still dictate what was proper in such cases. The general rule would still apply.

    If anything, it would be going in the other direction where an argument could have been made that the rule didn't apply. If the court had only ever ruled that it applied in cases involving FRAND-bound patents, then it might be argued whether it also applies in cases which didn't involve FRAND-bound patents. Indeed, that was one of the issues addressed in CSIRO v Cisco (Federal Circuit, 2015). Do the rules from Ericsson apply even if the patents in question don't carry FRAND obligations?

    But, to be clear: Ericsson v D-Link (2014) - a case which addressed a number of legal issues - involved patents that carried FRAND obligations. And in that case the Federal Circuit reiterated the entire market value rule which I've been referring to. It said, among other things (citations omitted):

    Thus, where the entire value of a machine as a marketable article is “properly and legally attributable to the patented feature,” the damages owed to the patentee may be calculated by reference to that value. Where it is not, however, courts must insist on a more realistic starting point for the royalty calculations by juries—often, the smallest salable unit and, at times, even less.
    The Federal Circuit has now reiterated this rule in a number of cases and contexts. And it follows from more general Supreme Court precedent that's more than a century old. That is the state of the law as it exists now. It could, of course, change in the future. 



    IMO following the decision in Unwired Planet v. Huawei (and confirmed on appeal) may make it a global licensing issue rather than one decided under US law. 
    "Just as implementers need protection, so too do the SEP owners. They are entitled to an appropriate reward for carrying out their research and development activities and for engaging with the standardisation process, and they must be able to prevent technology users from free-riding on their innovations. It is therefore important that implementers engage constructively in any FRAND negotiation and, where necessary, agree to submit to the outcome of an appropriate FRAND determination."

    "When a patent holder declares a patent essential to an ETSI telecommunications standard, it must make a commitment to license that patent on FRAND terms. In this judgment, the court has held that a FRAND licence will typically be worldwide and that the terms of that licence should not vary substantially depending on the size of the licensee/implementer...
     
    The court confirmed that both parties must take a FRAND approach; the SEP holder must not hold up (effectively refusing to license) and the licensee/implementer must not hold out (effectively refusing to take a licence).  However, this does not mean that all offers made during negotiations have to be FRAND. A FRAND approach still allows for ‘non-FRAND’ opening offers that leave room for negotiation upwards or downwards towards an eventual FRAND licence.  Offers above or below FRAND will only become problematic if they are extreme or take an intransigent approach which disrupts or prejudices the negotiation. Notably, in the judge’s view, it is not hold up for an SEP holder to seek to obtain through its licensing some of the value that results from the inclusion of its technology into a standard.

    As a result of a particular issue raised in this case, the court also dealt specifically with the ‘non-discrimination’ or ND part of FRAND.  He held that a licensee/implementer is not necessarily entitled to demand the same licence terms as another similarly positioned licensee (although it may be able to if the differences in licence terms distort competition between the two licensees).  Instead the ‘non-discrimination’ part of FRAND means that licensees should be offered FRAND royalties based upon the strength of the patent portfolio and not the size of the licensee."

    Qualcomm's patent portfolio would be considered strong would it not, and has demonstrated that it is willing to negotiate and therefor not "intransigent? Further, as far as the US is concerned, wouldn't the SEP licensing terms set out in the contractual agreement with ETSI that made the portfolio available to willing licensees on a FRAND basis carry extreme weight? Those royalities were agreed to be based on a percentage of the finished device cost, as was stated by contributors from LG to Motorola, to Huawei to Nokia, and ZTE to Qualcomm. While the specific percentage may be modified as part of negotiations between the parties I don't see how the companies who contractually agreed to the terms under which the IP would be contributed can now be found to be acting illegally if collecting royalties on the finished device. 

    I'm in no way saying I personally think it should remain unchanged, but it is what it is. Tossing out a standards contract before the end of its term will come with a new set of problems going forward. I think that's one reason among many why no US court has done so and won't. 

    So I don't see Qualcomm at this stage being globally barred from collecting royalties based on a completed device cost, particularly with the limitations (some new) they've agreed to in accordance with the standards organizations, ie tiered royalties and a reduced upper limit on the cost. Nor do I see any way that Apple can avoid paying royalties to Qualcomm on a per-device basis any time in the near future even if not using any Qualcomm hardware.  While highly unlikely in the US in at least three venues now, China, Germany, and the UK, unwillingness to pay royalties for the use of standards essential IP can be grounds for an injunction on sales of the offending product, an outcome Apple will avoid if push comes to shove. IMO IP holders being forced into negotiating different licensing rates and terms across various markets is also becoming less likely based on recent rulings. 


    edited November 2018
  • Reply 33 of 39
    carnegiecarnegie Posts: 1,077member
    gatorguy said:
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
    Apple won't necessarily have to pay Qualcomm royalties, based on the entire market value, when it doesn't use Qualcomm hardware (and even when it does, for that matter). That's part of the import of the ruling which is the subject of this thread.

    Apple might agree to pay royalties directly to Qualcomm on the entire market value, if the terms of its agreement with Qualcomm are otherwise favorable enough. But Apple has the a lot leverage now when it comes to whether it agrees to do so and on what terms it might agree to do so.

    One of the things which Qualcomm used to do, which violated its FRAND obligations, was refuse to license directly to its modem competitors. That allowed it to do some of the other improper things that it did. That's how Qualcomm's scheme worked, it employed a number of (illegal or contract violative) tactics which all worked together and which wouldn't have been effective on their own.

    If Qualcomm licensed to, e.g. Intel, directly, then Apple wouldn't need to pay Qualcomm for the IP in the modems which it buys from Intel. Qualcomm's IP rights would be exhausted. So it wouldn't have a way to charge Apple based on the entire market value (of end products). Whatever Intel pays wouldn't be based on the value of the devices its modems go into. At most, it would be based on the value of the modems (and even that wouldn't necessarily be the case, based on recent Federal Circuit decisions). This is why Qualcomm had previously refused to license to, e.g., Intel.

    When it comes to other Qualcomm IP which Apple might want to license - IP which isn't FRAND encumbered - Qualcomm's case for requiring royalties based on the entire market value has only gotten weaker. The Federal Circuit had already been clear that patent holders weren't entitled to royalties based on the entire market value unless certain conditions were met, and it's safe to say that those conditions weren't met with regard to, e.g., iPhones. But with its recent decision in Power Integrations v Fairchild Semiconductor International, the Federal Circuit arguably made it even more difficult for patent holders to establish that they were entitled to royalties based on the entire market value (citations omitted):

    As LaserDynamics, Versata, and VirnetX held, the entire market value rule is appropriate only when the patented feature is the sole driver of customer demand or substantially creates the value of the component parts. The burden of proof in this respect is on the patent holder. The question is whether the accused product, compared to other products in the same field, has features that would cause consumers to purchase the products beyond the patented feature, i.e., valuable features. Where the accused infringer presents evidence that its accused product has other valuable features beyond the patented feature, the patent holder must establish that these features do not cause consumers to purchase the product. A patentee may do this by showing that the patented feature “alone motivates customers to purchase [the infringing product]” in the first place. But when the product contains multiple valuable features, it is not enough to merely show that the patented feature is viewed as essential, that a product would not be commercially viable without the patented feature, or that consumers would not purchase the product without the patented feature. When the product contains other valuable features, the patentee must prove that those other features do not cause consumers to purchase the product. 
    You could well be correct a few years down the road, but with several lawsuits over Qualcomm licensing terms settled and others underway has any court yet said that royalties based on a total device cost is illegal or even un-FRAND-ly? 
    My comments were based on the state of the law as it is today, not as it might be in the future.

    As I've indicated before, royalties based on the entire market value of devices aren't illegal. Parties can freely agree to such terms. Apple could, e.g., decide that the rest of the terms of a comprehensive agreement with Qualcomm were favorable enough that it would be willing to pay royalties to Qualcomm based on the entire market value of its devices.

    But when it comes to what Qualcomm is entitled to - e.g., what it should be awarded by a court if it can't agree with Apple on what proper royalties should be - the Federal Circuit has been clear on this point. Patent holders are not entitled to royalties based on the entire market value unless they can demonstrate that their patents create all of the demand form the products in question, that other features of those products don't drive any of the demand for them. And when it comes to SEPs in particular, the Federal Circuit has also been clear in that the value that their holders' are entitled to is that intrinsic to the patented inventions - not that which is created by those inventions being included in the relevant standards.

    In deciding patent royalty cases, and setting out the rules which are to apply to damage calculations, the Federal Circuit is deciding what is reasonable in terms of what is owed to patent holders. FRAND terms, as you are aware, generally aren't specified in detail in the agreements which standard-setting organizations make with SEP holders. What represents and doesn't represent FRAND terms is determined by industry practices and case law. The latter, to the extent it exists on a particular point, is what controls.

    The Federal Circuit saying which methods are proper when it comes to calculating royalties owed to patent holders - in cases where parties don't agree - determines what is reasonable when it comes to SEPs bound by FRAND terms. Relevant here, even if the Federal Circuit hadn't applied its entire market value rule in cases that specifically dealt with FRAND-bound patents, that rule would still dictate what was proper in such cases. The general rule would still apply.

    If anything, it would be going in the other direction where an argument could have been made that the rule didn't apply. If the court had only ever ruled that it applied in cases involving FRAND-bound patents, then it might be argued whether it also applies in cases which didn't involve FRAND-bound patents. Indeed, that was one of the issues addressed in CSIRO v Cisco (Federal Circuit, 2015). Do the rules from Ericsson apply even if the patents in question don't carry FRAND obligations?

    But, to be clear: Ericsson v D-Link (2014) - a case which addressed a number of legal issues - involved patents that carried FRAND obligations. And in that case the Federal Circuit reiterated the entire market value rule which I've been referring to. It said, among other things (citations omitted):

    Thus, where the entire value of a machine as a marketable article is “properly and legally attributable to the patented feature,” the damages owed to the patentee may be calculated by reference to that value. Where it is not, however, courts must insist on a more realistic starting point for the royalty calculations by juries—often, the smallest salable unit and, at times, even less.
    The Federal Circuit has now reiterated this rule in a number of cases and contexts. And it follows from more general Supreme Court precedent that's more than a century old. That is the state of the law as it exists now. It could, of course, change in the future. 



    IMO following the decision in Unwired Planet v. Huawei (and confirmed on appeal) may make it a global licensing issue rather than one decided under US law. 
    "Just as implementers need protection, so too do the SEP owners. They are entitled to an appropriate reward for carrying out their research and development activities and for engaging with the standardisation process, and they must be able to prevent technology users from free-riding on their innovations. It is therefore important that implementers engage constructively in any FRAND negotiation and, where necessary, agree to submit to the outcome of an appropriate FRAND determination."

    "When a patent holder declares a patent essential to an ETSI telecommunications standard, it must make a commitment to license that patent on FRAND terms. In this judgment, the court has held that a FRAND licence will typically be worldwide and that the terms of that licence should not vary substantially depending on the size of the licensee/implementer...
     
    The court confirmed that both parties must take a FRAND approach; the SEP holder must not hold up (effectively refusing to license) and the licensee/implementer must not hold out (effectively refusing to take a licence).  However, this does not mean that all offers made during negotiations have to be FRAND. A FRAND approach still allows for ‘non-FRAND’ opening offers that leave room for negotiation upwards or downwards towards an eventual FRAND licence.  Offers above or below FRAND will only become problematic if they are extreme or take an intransigent approach which disrupts or prejudices the negotiation. Notably, in the judge’s view, it is not hold up for an SEP holder to seek to obtain through its licensing some of the value that results from the inclusion of its technology into a standard.

    As a result of a particular issue raised in this case, the court also dealt specifically with the ‘non-discrimination’ or ND part of FRAND.  He held that a licensee/implementer is not necessarily entitled to demand the same licence terms as another similarly positioned licensee (although it may be able to if the differences in licence terms distort competition between the two licensees).  Instead the ‘non-discrimination’ part of FRAND means that licensees should be offered FRAND royalties based upon the strength of the patent portfolio and not the size of the licensee."

    Qualcomm's patent portfolio would be considered strong would it not, and has demonstrated that it is willing to negotiate and therefor not "intransigent? Further, as far as the US is concerned, wouldn't the SEP licensing terms set out in the contractual agreement with ETSI that made the portfolio available to willing licensees on a FRAND basis carry extreme weight? Those royalities were agreed to be based on a percentage of the finished device cost, as was stated by contributors from LG to Motorola, to Huawei to Nokia, and ZTE to Qualcomm. While the specific percentage may be modified as part of negotiations between the parties I don't see how the companies who contractually agreed to the terms under which the IP would be contributed can now be found to be acting illegally if collecting royalties on the finished device. 

    I'm in no way saying I personally think it should remain unchanged, but it is what it is. Tossing out a standards contract before the end of its term will come with a new set of problems going forward. I think that's one reason among many why no US court has done so and won't. 

    So I don't see Qualcomm at this stage being globally barred from collecting royalties based on a completed device cost, particularly with the limitations (some new) they've agreed to in accordance with the standards organizations, ie tiered royalties and a reduced upper limit on the cost. Nor do I see any way that Apple can avoid paying royalties to Qualcomm on a per-device basis any time in the near future even if not using any Qualcomm hardware.  While highly unlikely in the US in at least three venues now, China, Germany, and the UK, unwillingness to pay royalties for the use of standards essential IP can be grounds for an injunction on sales of the offending product, an outcome Apple will avoid if push comes to shove. IMO IP holders being forced into negotiating different licensing rates and terms across various markets is also becoming less likely based on recent rulings. 


    A few things...

    (1) Again, no one is suggesting that Qualcomm collecting royalties based on end device cost is illegal. And no one is suggesting they'd be globally barred from doing so. Qualcomm and licensees are free to agree to such terms. The issue is whether Qualcomm can force would-be licensees to agree to such terms and, when it comes to this conversation in particular, whether Qualcomm would be able to get a court to order such terms.

    (2) Whether Qualcomm has acted in good faith in licensing negotiations doesn't really matter when it comes to the points I made (about Apple ending up having to pay royalties based on end product costs). But if the allegations made by a number of industry participants (e.g. Apple, Intel, Samsung) and the findings made by a number of regulatory bodies are correct, then it is fair to say that Qualcomm has been intransigent when it comes to licensing negotiations. It's been intransigent in multiple regards. That's part of the point when it comes to the actions brought by regulatory agencies and, e.g., Apple.

    (3) The decision you're referring to is from a U.K. court. It doesn't control U.S. law and the decisions of U.S. courts. U.S. courts will sometimes consider what foreign courts have had to say (depending, of course, on context), but they don't necessarily give it great weight. And, generally speaking, they aren't going to defer to what foreign courts have found on points of law if such points of law are already settled (to the contrary) for purposes of U.S. law. That ruling from the U.K. High Court of Justice is in clear conflict with U.S. law in some regards. So I wouldn't consider it particularly relevant when we're talking about what's likely to happen as a result of a case in U.S. courts between parties which are U.S. companies.

    (4) My original post referred to what might happen with regard to non-FRAND Qualcomm IP which Apple might want to license. The terms of the FRAND agreements Qualcomm has made wouldn't be particularly relevant in that regard.

    (5) When it comes to FRAND-bound IP, the terms of those agreements would of course be relevant. Is it your suggestion that the actual contractual terms Qualcomm agreed to specified that royalties based on end product cost would be considered FRAND? If so, I'd appreciate a pointer to those actual terms. That would be new information for me. Qualcomm hasn't even claimed that, as far as I'm aware, in its legal arguments. That would certainly matter. However, even if that were the case, it wouldn't mean that under U.S. law Qualcomm was entitled to such terms (if, e.g., Apple was unwilling to agree to them). It would just mean that such terms were allowed for FRAND purposes. U.S. law could, in effect, set further limits on what Qualcomm was entitled to, notwithstanding contractual commitments which Qualcomm made. And in this context Apple would be a third-party beneficiary of the agreements Qualcomm has made with SSOs, not an actual party to such agreements. So it likely wouldn't be contractually bound to such terms if U.S. law entitled it to more favorable terms.

    (6) To be clear, I'm not suggesting that Qualcomm and Apple will end up with different licensing terms for different parts of the world. The reality is that Qualcomm has, for all intents and purposes, already lost this fight. It's lost the leverage. It's been required to stop (in the U.S. and in other jurisdictions) doing many of the things it was doing before. And market developments have taken away its ability to do certain things. The scheme it had previously used relied on it doing a bunch of things which worked together to limit industry players' ability to get out from under its unilateral terms. That scheme just can't work anymore, too many pieces of the puzzle have been destroyed. So the question isn't whether Qualcomm gets to go on mostly doing what it had been doing. The question isn't can Qualcomm still win this fight. The question is how painful will its loss be. The question is in the details. What specifically will Qualcomm be left with when it comes to its ability to collect licensing fees? I expect that Apple and Qualcomm will eventually settle all of their ongoing legal fights and agree to comprehensive (global) terms. But those terms will be far more favorable to Apple than the terms which had previously been unilaterally imposed by Qualcomm. Will it involve per-device royalties? As I suggested previously... possibly. But I expect that will only happen if the terms are otherwise very favorable for Apple.

    I'd add that I'm not among those that thinks this dooms Qualcomm. I don't see it heading toward bankruptcy. I think it will be fine. It just won't be as profitable as it otherwise would have been going forward, and there will be considerably more competition in, e.g., the modem market.
  • Reply 34 of 39
    gatorguygatorguy Posts: 24,176member
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
    Apple won't necessarily have to pay Qualcomm royalties, based on the entire market value, when it doesn't use Qualcomm hardware (and even when it does, for that matter). That's part of the import of the ruling which is the subject of this thread.

    Apple might agree to pay royalties directly to Qualcomm on the entire market value, if the terms of its agreement with Qualcomm are otherwise favorable enough. But Apple has the a lot leverage now when it comes to whether it agrees to do so and on what terms it might agree to do so.

    One of the things which Qualcomm used to do, which violated its FRAND obligations, was refuse to license directly to its modem competitors. That allowed it to do some of the other improper things that it did. That's how Qualcomm's scheme worked, it employed a number of (illegal or contract violative) tactics which all worked together and which wouldn't have been effective on their own.

    If Qualcomm licensed to, e.g. Intel, directly, then Apple wouldn't need to pay Qualcomm for the IP in the modems which it buys from Intel. Qualcomm's IP rights would be exhausted. So it wouldn't have a way to charge Apple based on the entire market value (of end products). Whatever Intel pays wouldn't be based on the value of the devices its modems go into. At most, it would be based on the value of the modems (and even that wouldn't necessarily be the case, based on recent Federal Circuit decisions). This is why Qualcomm had previously refused to license to, e.g., Intel.

    When it comes to other Qualcomm IP which Apple might want to license - IP which isn't FRAND encumbered - Qualcomm's case for requiring royalties based on the entire market value has only gotten weaker. The Federal Circuit had already been clear that patent holders weren't entitled to royalties based on the entire market value unless certain conditions were met, and it's safe to say that those conditions weren't met with regard to, e.g., iPhones. But with its recent decision in Power Integrations v Fairchild Semiconductor International, the Federal Circuit arguably made it even more difficult for patent holders to establish that they were entitled to royalties based on the entire market value (citations omitted):

    As LaserDynamics, Versata, and VirnetX held, the entire market value rule is appropriate only when the patented feature is the sole driver of customer demand or substantially creates the value of the component parts. The burden of proof in this respect is on the patent holder. The question is whether the accused product, compared to other products in the same field, has features that would cause consumers to purchase the products beyond the patented feature, i.e., valuable features. Where the accused infringer presents evidence that its accused product has other valuable features beyond the patented feature, the patent holder must establish that these features do not cause consumers to purchase the product. A patentee may do this by showing that the patented feature “alone motivates customers to purchase [the infringing product]” in the first place. But when the product contains multiple valuable features, it is not enough to merely show that the patented feature is viewed as essential, that a product would not be commercially viable without the patented feature, or that consumers would not purchase the product without the patented feature. When the product contains other valuable features, the patentee must prove that those other features do not cause consumers to purchase the product. 
    You could well be correct a few years down the road, but with several lawsuits over Qualcomm licensing terms settled and others underway has any court yet said that royalties based on a total device cost is illegal or even un-FRAND-ly? 
    My comments were based on the state of the law as it is today, not as it might be in the future.

    As I've indicated before, royalties based on the entire market value of devices aren't illegal. Parties can freely agree to such terms. Apple could, e.g., decide that the rest of the terms of a comprehensive agreement with Qualcomm were favorable enough that it would be willing to pay royalties to Qualcomm based on the entire market value of its devices.

    But when it comes to what Qualcomm is entitled to - e.g., what it should be awarded by a court if it can't agree with Apple on what proper royalties should be - the Federal Circuit has been clear on this point. Patent holders are not entitled to royalties based on the entire market value unless they can demonstrate that their patents create all of the demand form the products in question, that other features of those products don't drive any of the demand for them. And when it comes to SEPs in particular, the Federal Circuit has also been clear in that the value that their holders' are entitled to is that intrinsic to the patented inventions - not that which is created by those inventions being included in the relevant standards.

    In deciding patent royalty cases, and setting out the rules which are to apply to damage calculations, the Federal Circuit is deciding what is reasonable in terms of what is owed to patent holders. FRAND terms, as you are aware, generally aren't specified in detail in the agreements which standard-setting organizations make with SEP holders. What represents and doesn't represent FRAND terms is determined by industry practices and case law. The latter, to the extent it exists on a particular point, is what controls.

    The Federal Circuit saying which methods are proper when it comes to calculating royalties owed to patent holders - in cases where parties don't agree - determines what is reasonable when it comes to SEPs bound by FRAND terms. Relevant here, even if the Federal Circuit hadn't applied its entire market value rule in cases that specifically dealt with FRAND-bound patents, that rule would still dictate what was proper in such cases. The general rule would still apply.

    If anything, it would be going in the other direction where an argument could have been made that the rule didn't apply. If the court had only ever ruled that it applied in cases involving FRAND-bound patents, then it might be argued whether it also applies in cases which didn't involve FRAND-bound patents. Indeed, that was one of the issues addressed in CSIRO v Cisco (Federal Circuit, 2015). Do the rules from Ericsson apply even if the patents in question don't carry FRAND obligations?

    But, to be clear: Ericsson v D-Link (2014) - a case which addressed a number of legal issues - involved patents that carried FRAND obligations. And in that case the Federal Circuit reiterated the entire market value rule which I've been referring to. It said, among other things (citations omitted):

    Thus, where the entire value of a machine as a marketable article is “properly and legally attributable to the patented feature,” the damages owed to the patentee may be calculated by reference to that value. Where it is not, however, courts must insist on a more realistic starting point for the royalty calculations by juries—often, the smallest salable unit and, at times, even less.
    The Federal Circuit has now reiterated this rule in a number of cases and contexts. And it follows from more general Supreme Court precedent that's more than a century old. That is the state of the law as it exists now. It could, of course, change in the future. 



    IMO following the decision in Unwired Planet v. Huawei (and confirmed on appeal) may make it a global licensing issue rather than one decided under US law. 
    "Just as implementers need protection, so too do the SEP owners. They are entitled to an appropriate reward for carrying out their research and development activities and for engaging with the standardisation process, and they must be able to prevent technology users from free-riding on their innovations. It is therefore important that implementers engage constructively in any FRAND negotiation and, where necessary, agree to submit to the outcome of an appropriate FRAND determination."

    "When a patent holder declares a patent essential to an ETSI telecommunications standard, it must make a commitment to license that patent on FRAND terms. In this judgment, the court has held that a FRAND licence will typically be worldwide and that the terms of that licence should not vary substantially depending on the size of the licensee/implementer...
     
    The court confirmed that both parties must take a FRAND approach; the SEP holder must not hold up (effectively refusing to license) and the licensee/implementer must not hold out (effectively refusing to take a licence).  However, this does not mean that all offers made during negotiations have to be FRAND. A FRAND approach still allows for ‘non-FRAND’ opening offers that leave room for negotiation upwards or downwards towards an eventual FRAND licence.  Offers above or below FRAND will only become problematic if they are extreme or take an intransigent approach which disrupts or prejudices the negotiation. Notably, in the judge’s view, it is not hold up for an SEP holder to seek to obtain through its licensing some of the value that results from the inclusion of its technology into a standard.

    As a result of a particular issue raised in this case, the court also dealt specifically with the ‘non-discrimination’ or ND part of FRAND.  He held that a licensee/implementer is not necessarily entitled to demand the same licence terms as another similarly positioned licensee (although it may be able to if the differences in licence terms distort competition between the two licensees).  Instead the ‘non-discrimination’ part of FRAND means that licensees should be offered FRAND royalties based upon the strength of the patent portfolio and not the size of the licensee."

    Qualcomm's patent portfolio would be considered strong would it not, and has demonstrated that it is willing to negotiate and therefor not "intransigent? Further, as far as the US is concerned, wouldn't the SEP licensing terms set out in the contractual agreement with ETSI that made the portfolio available to willing licensees on a FRAND basis carry extreme weight? Those royalities were agreed to be based on a percentage of the finished device cost, as was stated by contributors from LG to Motorola, to Huawei to Nokia, and ZTE to Qualcomm. While the specific percentage may be modified as part of negotiations between the parties I don't see how the companies who contractually agreed to the terms under which the IP would be contributed can now be found to be acting illegally if collecting royalties on the finished device. 

    I'm in no way saying I personally think it should remain unchanged, but it is what it is. Tossing out a standards contract before the end of its term will come with a new set of problems going forward. I think that's one reason among many why no US court has done so and won't. 

    So I don't see Qualcomm at this stage being globally barred from collecting royalties based on a completed device cost, particularly with the limitations (some new) they've agreed to in accordance with the standards organizations, ie tiered royalties and a reduced upper limit on the cost. Nor do I see any way that Apple can avoid paying royalties to Qualcomm on a per-device basis any time in the near future even if not using any Qualcomm hardware.  While highly unlikely in the US in at least three venues now, China, Germany, and the UK, unwillingness to pay royalties for the use of standards essential IP can be grounds for an injunction on sales of the offending product, an outcome Apple will avoid if push comes to shove. IMO IP holders being forced into negotiating different licensing rates and terms across various markets is also becoming less likely based on recent rulings. 


    A few things...

    (1) Again, no one is suggesting that Qualcomm collecting royalties based on end device cost is illegal. 
    That was exactly the claim made by several posters in this thread which is what led into the discussion you and I are having. :)
     ...and yes, you have several valid points regarding recent court decisions, but to date those have all been case-specific involving individual contracts unique to particular parties, not FRAND licensing in general, the Unwired/Huawei ruling being an exception IMO. Many end up being at opposition if delving into the details, the TCL v. Ericsson one you brought up and Unwired Planet v. Huawei that I referred to for example.

    Still as you say there's much yet to be ruled on before there's clarity in the matter of SEP licensing. There's not even a definitive "what is FRAND" yet. By the way there's an excellent interview over at IAM involving multiple IP attorneys with expertise in SEP commenting on some of the recent cases which you might have an interest in. https://www.iam-media.com/frandseps/sep-licensing-standard

    As always I thoroughly appreciate your posts on the subject. Thanks! Great having knowledgeably folks involved in discussions, particularly those like you who take time to try to understand the topics. 
    edited November 2018
  • Reply 35 of 39
    carnegiecarnegie Posts: 1,077member
    gatorguy said:
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
    Apple won't necessarily have to pay Qualcomm royalties, based on the entire market value, when it doesn't use Qualcomm hardware (and even when it does, for that matter). That's part of the import of the ruling which is the subject of this thread.

    Apple might agree to pay royalties directly to Qualcomm on the entire market value, if the terms of its agreement with Qualcomm are otherwise favorable enough. But Apple has the a lot leverage now when it comes to whether it agrees to do so and on what terms it might agree to do so.

    One of the things which Qualcomm used to do, which violated its FRAND obligations, was refuse to license directly to its modem competitors. That allowed it to do some of the other improper things that it did. That's how Qualcomm's scheme worked, it employed a number of (illegal or contract violative) tactics which all worked together and which wouldn't have been effective on their own.

    If Qualcomm licensed to, e.g. Intel, directly, then Apple wouldn't need to pay Qualcomm for the IP in the modems which it buys from Intel. Qualcomm's IP rights would be exhausted. So it wouldn't have a way to charge Apple based on the entire market value (of end products). Whatever Intel pays wouldn't be based on the value of the devices its modems go into. At most, it would be based on the value of the modems (and even that wouldn't necessarily be the case, based on recent Federal Circuit decisions). This is why Qualcomm had previously refused to license to, e.g., Intel.

    When it comes to other Qualcomm IP which Apple might want to license - IP which isn't FRAND encumbered - Qualcomm's case for requiring royalties based on the entire market value has only gotten weaker. The Federal Circuit had already been clear that patent holders weren't entitled to royalties based on the entire market value unless certain conditions were met, and it's safe to say that those conditions weren't met with regard to, e.g., iPhones. But with its recent decision in Power Integrations v Fairchild Semiconductor International, the Federal Circuit arguably made it even more difficult for patent holders to establish that they were entitled to royalties based on the entire market value (citations omitted):

    As LaserDynamics, Versata, and VirnetX held, the entire market value rule is appropriate only when the patented feature is the sole driver of customer demand or substantially creates the value of the component parts. The burden of proof in this respect is on the patent holder. The question is whether the accused product, compared to other products in the same field, has features that would cause consumers to purchase the products beyond the patented feature, i.e., valuable features. Where the accused infringer presents evidence that its accused product has other valuable features beyond the patented feature, the patent holder must establish that these features do not cause consumers to purchase the product. A patentee may do this by showing that the patented feature “alone motivates customers to purchase [the infringing product]” in the first place. But when the product contains multiple valuable features, it is not enough to merely show that the patented feature is viewed as essential, that a product would not be commercially viable without the patented feature, or that consumers would not purchase the product without the patented feature. When the product contains other valuable features, the patentee must prove that those other features do not cause consumers to purchase the product. 
    You could well be correct a few years down the road, but with several lawsuits over Qualcomm licensing terms settled and others underway has any court yet said that royalties based on a total device cost is illegal or even un-FRAND-ly? 
    My comments were based on the state of the law as it is today, not as it might be in the future.

    As I've indicated before, royalties based on the entire market value of devices aren't illegal. Parties can freely agree to such terms. Apple could, e.g., decide that the rest of the terms of a comprehensive agreement with Qualcomm were favorable enough that it would be willing to pay royalties to Qualcomm based on the entire market value of its devices.

    But when it comes to what Qualcomm is entitled to - e.g., what it should be awarded by a court if it can't agree with Apple on what proper royalties should be - the Federal Circuit has been clear on this point. Patent holders are not entitled to royalties based on the entire market value unless they can demonstrate that their patents create all of the demand form the products in question, that other features of those products don't drive any of the demand for them. And when it comes to SEPs in particular, the Federal Circuit has also been clear in that the value that their holders' are entitled to is that intrinsic to the patented inventions - not that which is created by those inventions being included in the relevant standards.

    In deciding patent royalty cases, and setting out the rules which are to apply to damage calculations, the Federal Circuit is deciding what is reasonable in terms of what is owed to patent holders. FRAND terms, as you are aware, generally aren't specified in detail in the agreements which standard-setting organizations make with SEP holders. What represents and doesn't represent FRAND terms is determined by industry practices and case law. The latter, to the extent it exists on a particular point, is what controls.

    The Federal Circuit saying which methods are proper when it comes to calculating royalties owed to patent holders - in cases where parties don't agree - determines what is reasonable when it comes to SEPs bound by FRAND terms. Relevant here, even if the Federal Circuit hadn't applied its entire market value rule in cases that specifically dealt with FRAND-bound patents, that rule would still dictate what was proper in such cases. The general rule would still apply.

    If anything, it would be going in the other direction where an argument could have been made that the rule didn't apply. If the court had only ever ruled that it applied in cases involving FRAND-bound patents, then it might be argued whether it also applies in cases which didn't involve FRAND-bound patents. Indeed, that was one of the issues addressed in CSIRO v Cisco (Federal Circuit, 2015). Do the rules from Ericsson apply even if the patents in question don't carry FRAND obligations?

    But, to be clear: Ericsson v D-Link (2014) - a case which addressed a number of legal issues - involved patents that carried FRAND obligations. And in that case the Federal Circuit reiterated the entire market value rule which I've been referring to. It said, among other things (citations omitted):

    Thus, where the entire value of a machine as a marketable article is “properly and legally attributable to the patented feature,” the damages owed to the patentee may be calculated by reference to that value. Where it is not, however, courts must insist on a more realistic starting point for the royalty calculations by juries—often, the smallest salable unit and, at times, even less.
    The Federal Circuit has now reiterated this rule in a number of cases and contexts. And it follows from more general Supreme Court precedent that's more than a century old. That is the state of the law as it exists now. It could, of course, change in the future. 



    IMO following the decision in Unwired Planet v. Huawei (and confirmed on appeal) may make it a global licensing issue rather than one decided under US law. 
    "Just as implementers need protection, so too do the SEP owners. They are entitled to an appropriate reward for carrying out their research and development activities and for engaging with the standardisation process, and they must be able to prevent technology users from free-riding on their innovations. It is therefore important that implementers engage constructively in any FRAND negotiation and, where necessary, agree to submit to the outcome of an appropriate FRAND determination."

    "When a patent holder declares a patent essential to an ETSI telecommunications standard, it must make a commitment to license that patent on FRAND terms. In this judgment, the court has held that a FRAND licence will typically be worldwide and that the terms of that licence should not vary substantially depending on the size of the licensee/implementer...
     
    The court confirmed that both parties must take a FRAND approach; the SEP holder must not hold up (effectively refusing to license) and the licensee/implementer must not hold out (effectively refusing to take a licence).  However, this does not mean that all offers made during negotiations have to be FRAND. A FRAND approach still allows for ‘non-FRAND’ opening offers that leave room for negotiation upwards or downwards towards an eventual FRAND licence.  Offers above or below FRAND will only become problematic if they are extreme or take an intransigent approach which disrupts or prejudices the negotiation. Notably, in the judge’s view, it is not hold up for an SEP holder to seek to obtain through its licensing some of the value that results from the inclusion of its technology into a standard.

    As a result of a particular issue raised in this case, the court also dealt specifically with the ‘non-discrimination’ or ND part of FRAND.  He held that a licensee/implementer is not necessarily entitled to demand the same licence terms as another similarly positioned licensee (although it may be able to if the differences in licence terms distort competition between the two licensees).  Instead the ‘non-discrimination’ part of FRAND means that licensees should be offered FRAND royalties based upon the strength of the patent portfolio and not the size of the licensee."

    Qualcomm's patent portfolio would be considered strong would it not, and has demonstrated that it is willing to negotiate and therefor not "intransigent? Further, as far as the US is concerned, wouldn't the SEP licensing terms set out in the contractual agreement with ETSI that made the portfolio available to willing licensees on a FRAND basis carry extreme weight? Those royalities were agreed to be based on a percentage of the finished device cost, as was stated by contributors from LG to Motorola, to Huawei to Nokia, and ZTE to Qualcomm. While the specific percentage may be modified as part of negotiations between the parties I don't see how the companies who contractually agreed to the terms under which the IP would be contributed can now be found to be acting illegally if collecting royalties on the finished device. 

    I'm in no way saying I personally think it should remain unchanged, but it is what it is. Tossing out a standards contract before the end of its term will come with a new set of problems going forward. I think that's one reason among many why no US court has done so and won't. 

    So I don't see Qualcomm at this stage being globally barred from collecting royalties based on a completed device cost, particularly with the limitations (some new) they've agreed to in accordance with the standards organizations, ie tiered royalties and a reduced upper limit on the cost. Nor do I see any way that Apple can avoid paying royalties to Qualcomm on a per-device basis any time in the near future even if not using any Qualcomm hardware.  While highly unlikely in the US in at least three venues now, China, Germany, and the UK, unwillingness to pay royalties for the use of standards essential IP can be grounds for an injunction on sales of the offending product, an outcome Apple will avoid if push comes to shove. IMO IP holders being forced into negotiating different licensing rates and terms across various markets is also becoming less likely based on recent rulings. 


    A few things...

    (1) Again, no one is suggesting that Qualcomm collecting royalties based on end device cost is illegal. 
    That was exactly the claim made by several posters in this thread which is what led into the discussion you and I are having. :)
     ...and yes, you have several valid points regarding recent court decisions, but to date those have all been case-specific involving individual contracts unique to particular parties, not FRAND licensing in general, the Unwired/Huawei ruling being an exception IMO. Many end up being at opposition if delving into the details, the TCL v. Ericsson one you brought up and Unwired Planet v. Huawei that I referred to for example.

    Still as you say there's much yet to be ruled on before there's clarity in the matter of SEP licensing. There's not even a definitive "what is FRAND" yet. By the way there's an excellent interview over at IAM involving multiple IP attorneys with expertise in SEP commenting on some of the recent cases which you might have an interest in. https://www.iam-media.com/frandseps/sep-licensing-standard

    As always I thoroughly appreciate your posts on the subject. Thanks! Great having knowledgeably folks involved in discussions, particularly those like you who take time to try to understand the topics. 
    Fair enough. If others have made such a claim - i.e. that collecting royalties based on end device cost is illegal - then that's not accurate. But I just skimmed the thread and didn't see such a claim. I see claims which I interpret to be more or less what I've said, that patent holders aren't entitled to such royalties if other parties aren't willing to agree to them. If both parties want to willingly agree to such terms, they can do so.

    As for court decisions being case-specific... yeah, each decision is specific to the case it is deciding. But they establish (or follow) legal principles and rules which are applicable to other cases. And the rule which I've referred to - the entire market value rule (which represents an exception to a broader principle) - is one that applies generally. It isn't limited in applicability to the specific cases I've referred to.
  • Reply 36 of 39
    gatorguygatorguy Posts: 24,176member
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
    Apple won't necessarily have to pay Qualcomm royalties, based on the entire market value, when it doesn't use Qualcomm hardware (and even when it does, for that matter). That's part of the import of the ruling which is the subject of this thread.

    Apple might agree to pay royalties directly to Qualcomm on the entire market value, if the terms of its agreement with Qualcomm are otherwise favorable enough. But Apple has the a lot leverage now when it comes to whether it agrees to do so and on what terms it might agree to do so.

    One of the things which Qualcomm used to do, which violated its FRAND obligations, was refuse to license directly to its modem competitors. That allowed it to do some of the other improper things that it did. That's how Qualcomm's scheme worked, it employed a number of (illegal or contract violative) tactics which all worked together and which wouldn't have been effective on their own.

    If Qualcomm licensed to, e.g. Intel, directly, then Apple wouldn't need to pay Qualcomm for the IP in the modems which it buys from Intel. Qualcomm's IP rights would be exhausted. So it wouldn't have a way to charge Apple based on the entire market value (of end products). Whatever Intel pays wouldn't be based on the value of the devices its modems go into. At most, it would be based on the value of the modems (and even that wouldn't necessarily be the case, based on recent Federal Circuit decisions). This is why Qualcomm had previously refused to license to, e.g., Intel.

    When it comes to other Qualcomm IP which Apple might want to license - IP which isn't FRAND encumbered - Qualcomm's case for requiring royalties based on the entire market value has only gotten weaker. The Federal Circuit had already been clear that patent holders weren't entitled to royalties based on the entire market value unless certain conditions were met, and it's safe to say that those conditions weren't met with regard to, e.g., iPhones. But with its recent decision in Power Integrations v Fairchild Semiconductor International, the Federal Circuit arguably made it even more difficult for patent holders to establish that they were entitled to royalties based on the entire market value (citations omitted):

    As LaserDynamics, Versata, and VirnetX held, the entire market value rule is appropriate only when the patented feature is the sole driver of customer demand or substantially creates the value of the component parts. The burden of proof in this respect is on the patent holder. The question is whether the accused product, compared to other products in the same field, has features that would cause consumers to purchase the products beyond the patented feature, i.e., valuable features. Where the accused infringer presents evidence that its accused product has other valuable features beyond the patented feature, the patent holder must establish that these features do not cause consumers to purchase the product. A patentee may do this by showing that the patented feature “alone motivates customers to purchase [the infringing product]” in the first place. But when the product contains multiple valuable features, it is not enough to merely show that the patented feature is viewed as essential, that a product would not be commercially viable without the patented feature, or that consumers would not purchase the product without the patented feature. When the product contains other valuable features, the patentee must prove that those other features do not cause consumers to purchase the product. 
    You could well be correct a few years down the road, but with several lawsuits over Qualcomm licensing terms settled and others underway has any court yet said that royalties based on a total device cost is illegal or even un-FRAND-ly? 
    My comments were based on the state of the law as it is today, not as it might be in the future.

    As I've indicated before, royalties based on the entire market value of devices aren't illegal. Parties can freely agree to such terms. Apple could, e.g., decide that the rest of the terms of a comprehensive agreement with Qualcomm were favorable enough that it would be willing to pay royalties to Qualcomm based on the entire market value of its devices.

    But when it comes to what Qualcomm is entitled to - e.g., what it should be awarded by a court if it can't agree with Apple on what proper royalties should be - the Federal Circuit has been clear on this point. Patent holders are not entitled to royalties based on the entire market value unless they can demonstrate that their patents create all of the demand form the products in question, that other features of those products don't drive any of the demand for them. And when it comes to SEPs in particular, the Federal Circuit has also been clear in that the value that their holders' are entitled to is that intrinsic to the patented inventions - not that which is created by those inventions being included in the relevant standards.

    In deciding patent royalty cases, and setting out the rules which are to apply to damage calculations, the Federal Circuit is deciding what is reasonable in terms of what is owed to patent holders. FRAND terms, as you are aware, generally aren't specified in detail in the agreements which standard-setting organizations make with SEP holders. What represents and doesn't represent FRAND terms is determined by industry practices and case law. The latter, to the extent it exists on a particular point, is what controls.

    The Federal Circuit saying which methods are proper when it comes to calculating royalties owed to patent holders - in cases where parties don't agree - determines what is reasonable when it comes to SEPs bound by FRAND terms. Relevant here, even if the Federal Circuit hadn't applied its entire market value rule in cases that specifically dealt with FRAND-bound patents, that rule would still dictate what was proper in such cases. The general rule would still apply.

    If anything, it would be going in the other direction where an argument could have been made that the rule didn't apply. If the court had only ever ruled that it applied in cases involving FRAND-bound patents, then it might be argued whether it also applies in cases which didn't involve FRAND-bound patents. Indeed, that was one of the issues addressed in CSIRO v Cisco (Federal Circuit, 2015). Do the rules from Ericsson apply even if the patents in question don't carry FRAND obligations?

    But, to be clear: Ericsson v D-Link (2014) - a case which addressed a number of legal issues - involved patents that carried FRAND obligations. And in that case the Federal Circuit reiterated the entire market value rule which I've been referring to. It said, among other things (citations omitted):

    Thus, where the entire value of a machine as a marketable article is “properly and legally attributable to the patented feature,” the damages owed to the patentee may be calculated by reference to that value. Where it is not, however, courts must insist on a more realistic starting point for the royalty calculations by juries—often, the smallest salable unit and, at times, even less.
    The Federal Circuit has now reiterated this rule in a number of cases and contexts. And it follows from more general Supreme Court precedent that's more than a century old. That is the state of the law as it exists now. It could, of course, change in the future. 



    IMO following the decision in Unwired Planet v. Huawei (and confirmed on appeal) may make it a global licensing issue rather than one decided under US law. 
    "Just as implementers need protection, so too do the SEP owners. They are entitled to an appropriate reward for carrying out their research and development activities and for engaging with the standardisation process, and they must be able to prevent technology users from free-riding on their innovations. It is therefore important that implementers engage constructively in any FRAND negotiation and, where necessary, agree to submit to the outcome of an appropriate FRAND determination."

    "When a patent holder declares a patent essential to an ETSI telecommunications standard, it must make a commitment to license that patent on FRAND terms. In this judgment, the court has held that a FRAND licence will typically be worldwide and that the terms of that licence should not vary substantially depending on the size of the licensee/implementer...
     
    The court confirmed that both parties must take a FRAND approach; the SEP holder must not hold up (effectively refusing to license) and the licensee/implementer must not hold out (effectively refusing to take a licence).  However, this does not mean that all offers made during negotiations have to be FRAND. A FRAND approach still allows for ‘non-FRAND’ opening offers that leave room for negotiation upwards or downwards towards an eventual FRAND licence.  Offers above or below FRAND will only become problematic if they are extreme or take an intransigent approach which disrupts or prejudices the negotiation. Notably, in the judge’s view, it is not hold up for an SEP holder to seek to obtain through its licensing some of the value that results from the inclusion of its technology into a standard.

    As a result of a particular issue raised in this case, the court also dealt specifically with the ‘non-discrimination’ or ND part of FRAND.  He held that a licensee/implementer is not necessarily entitled to demand the same licence terms as another similarly positioned licensee (although it may be able to if the differences in licence terms distort competition between the two licensees).  Instead the ‘non-discrimination’ part of FRAND means that licensees should be offered FRAND royalties based upon the strength of the patent portfolio and not the size of the licensee."

    Qualcomm's patent portfolio would be considered strong would it not, and has demonstrated that it is willing to negotiate and therefor not "intransigent? Further, as far as the US is concerned, wouldn't the SEP licensing terms set out in the contractual agreement with ETSI that made the portfolio available to willing licensees on a FRAND basis carry extreme weight? Those royalities were agreed to be based on a percentage of the finished device cost, as was stated by contributors from LG to Motorola, to Huawei to Nokia, and ZTE to Qualcomm. While the specific percentage may be modified as part of negotiations between the parties I don't see how the companies who contractually agreed to the terms under which the IP would be contributed can now be found to be acting illegally if collecting royalties on the finished device. 

    I'm in no way saying I personally think it should remain unchanged, but it is what it is. Tossing out a standards contract before the end of its term will come with a new set of problems going forward. I think that's one reason among many why no US court has done so and won't. 

    So I don't see Qualcomm at this stage being globally barred from collecting royalties based on a completed device cost, particularly with the limitations (some new) they've agreed to in accordance with the standards organizations, ie tiered royalties and a reduced upper limit on the cost. Nor do I see any way that Apple can avoid paying royalties to Qualcomm on a per-device basis any time in the near future even if not using any Qualcomm hardware.  While highly unlikely in the US in at least three venues now, China, Germany, and the UK, unwillingness to pay royalties for the use of standards essential IP can be grounds for an injunction on sales of the offending product, an outcome Apple will avoid if push comes to shove. IMO IP holders being forced into negotiating different licensing rates and terms across various markets is also becoming less likely based on recent rulings. 


    A few things...

    (1) Again, no one is suggesting that Qualcomm collecting royalties based on end device cost is illegal. 
    That was exactly the claim made by several posters in this thread which is what led into the discussion you and I are having. :)
     ...and yes, you have several valid points regarding recent court decisions, but to date those have all been case-specific involving individual contracts unique to particular parties, not FRAND licensing in general, the Unwired/Huawei ruling being an exception IMO. Many end up being at opposition if delving into the details, the TCL v. Ericsson one you brought up and Unwired Planet v. Huawei that I referred to for example.

    Still as you say there's much yet to be ruled on before there's clarity in the matter of SEP licensing. There's not even a definitive "what is FRAND" yet. By the way there's an excellent interview over at IAM involving multiple IP attorneys with expertise in SEP commenting on some of the recent cases which you might have an interest in. https://www.iam-media.com/frandseps/sep-licensing-standard

    As always I thoroughly appreciate your posts on the subject. Thanks! Great having knowledgeably folks involved in discussions, particularly those like you who take time to try to understand the topics. 
    As for court decisions being case-specific... yeah, each decision is specific to the case it is deciding. But they establish (or follow) legal principles and rules which are applicable to other cases. And the rule which I've referred to - the entire market value rule (which represents an exception to a broader principle) - is one that applies generally. It isn't limited in applicability to the specific cases I've referred to.
    Things are not as many here seem to believe they are when it comes to smallest saleable unit arguments as the SEP royalty basis. Even very recent Federal Court rulings have reiterated that royalties based on the end-product price rather than a component can be perfectly proper and thus legal, and on grounds beyond those you brought up in the
     Power Integrations v Fairchild Semiconductor International case.

    "The Federal Circuit’s recent Exmark v. Briggs-Stratton decision further confirms that there is no categorical rule about selecting a royalty base when litigating a reasonable royalty in order to apportion value to the patented invention, but that “apportionment can be addressed in a variety of ways … [s]o long as [the patent owner] adequately and reliably apportions between the improved and conventional features” of the accused product.  Thus, in this case, the Federal Circuit ruled that the patent owner properly could use the entire lawn mower as the royalty base and was not limited to the innovative baffle component of the lawn mower as a royalty base.  This case continues the clarification made in the Federal Circuit’s CSIRO decision involving standard essential patents that recognized that “adaptability [in determining patent damages] is necessary because different cases present different facts” and rejected as “untenable” the argument that every damages model must start with the smallest salable patent practicing unit"
    https://www.essentialpatentblog.com/2018/01/federal-circuit-confirms-flexibility-determining-royalty-base-exmark-v-briggs-stratton/#more-7848
    edited December 2018
  • Reply 37 of 39
    carnegiecarnegie Posts: 1,077member
    gatorguy said:
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    carnegie said:
    gatorguy said:
    chasm said:
    Regarding FRAND terms: there are plenty of models for similar tech deals and technology that Qualcomm should be made to follow
    See post 23 for tech companies that similarly charge royalties on the finished product price. There's no reason to think that once the dust settles on this set of lawsuits that Qualcomm will be barred from continuing to charge royalties based on a final build cost. They're not simply selling a chip but an entire package of IP essential to building a smartphone. Apple will still have to pay Qualcomm royalties in one way or another even if they don't use any of their hardware. It's all the peripheral stuff that's going to be reined in, if not now then by some other court or agency in the near future IMO. 

    chasm said:
    The overall point, though, is that Qualcomm -- whether it admits it or not -- has a business model that requires it to go both outside FRAND guidelines and outside the boundaries of a legit business model.
    Yes they have. Profit is fine. Greedy profit can cross the line into illegal. 
            
    Apple won't necessarily have to pay Qualcomm royalties, based on the entire market value, when it doesn't use Qualcomm hardware (and even when it does, for that matter). That's part of the import of the ruling which is the subject of this thread.

    Apple might agree to pay royalties directly to Qualcomm on the entire market value, if the terms of its agreement with Qualcomm are otherwise favorable enough. But Apple has the a lot leverage now when it comes to whether it agrees to do so and on what terms it might agree to do so.

    One of the things which Qualcomm used to do, which violated its FRAND obligations, was refuse to license directly to its modem competitors. That allowed it to do some of the other improper things that it did. That's how Qualcomm's scheme worked, it employed a number of (illegal or contract violative) tactics which all worked together and which wouldn't have been effective on their own.

    If Qualcomm licensed to, e.g. Intel, directly, then Apple wouldn't need to pay Qualcomm for the IP in the modems which it buys from Intel. Qualcomm's IP rights would be exhausted. So it wouldn't have a way to charge Apple based on the entire market value (of end products). Whatever Intel pays wouldn't be based on the value of the devices its modems go into. At most, it would be based on the value of the modems (and even that wouldn't necessarily be the case, based on recent Federal Circuit decisions). This is why Qualcomm had previously refused to license to, e.g., Intel.

    When it comes to other Qualcomm IP which Apple might want to license - IP which isn't FRAND encumbered - Qualcomm's case for requiring royalties based on the entire market value has only gotten weaker. The Federal Circuit had already been clear that patent holders weren't entitled to royalties based on the entire market value unless certain conditions were met, and it's safe to say that those conditions weren't met with regard to, e.g., iPhones. But with its recent decision in Power Integrations v Fairchild Semiconductor International, the Federal Circuit arguably made it even more difficult for patent holders to establish that they were entitled to royalties based on the entire market value (citations omitted):

    As LaserDynamics, Versata, and VirnetX held, the entire market value rule is appropriate only when the patented feature is the sole driver of customer demand or substantially creates the value of the component parts. The burden of proof in this respect is on the patent holder. The question is whether the accused product, compared to other products in the same field, has features that would cause consumers to purchase the products beyond the patented feature, i.e., valuable features. Where the accused infringer presents evidence that its accused product has other valuable features beyond the patented feature, the patent holder must establish that these features do not cause consumers to purchase the product. A patentee may do this by showing that the patented feature “alone motivates customers to purchase [the infringing product]” in the first place. But when the product contains multiple valuable features, it is not enough to merely show that the patented feature is viewed as essential, that a product would not be commercially viable without the patented feature, or that consumers would not purchase the product without the patented feature. When the product contains other valuable features, the patentee must prove that those other features do not cause consumers to purchase the product. 
    You could well be correct a few years down the road, but with several lawsuits over Qualcomm licensing terms settled and others underway has any court yet said that royalties based on a total device cost is illegal or even un-FRAND-ly? 
    My comments were based on the state of the law as it is today, not as it might be in the future.

    As I've indicated before, royalties based on the entire market value of devices aren't illegal. Parties can freely agree to such terms. Apple could, e.g., decide that the rest of the terms of a comprehensive agreement with Qualcomm were favorable enough that it would be willing to pay royalties to Qualcomm based on the entire market value of its devices.

    But when it comes to what Qualcomm is entitled to - e.g., what it should be awarded by a court if it can't agree with Apple on what proper royalties should be - the Federal Circuit has been clear on this point. Patent holders are not entitled to royalties based on the entire market value unless they can demonstrate that their patents create all of the demand form the products in question, that other features of those products don't drive any of the demand for them. And when it comes to SEPs in particular, the Federal Circuit has also been clear in that the value that their holders' are entitled to is that intrinsic to the patented inventions - not that which is created by those inventions being included in the relevant standards.

    In deciding patent royalty cases, and setting out the rules which are to apply to damage calculations, the Federal Circuit is deciding what is reasonable in terms of what is owed to patent holders. FRAND terms, as you are aware, generally aren't specified in detail in the agreements which standard-setting organizations make with SEP holders. What represents and doesn't represent FRAND terms is determined by industry practices and case law. The latter, to the extent it exists on a particular point, is what controls.

    The Federal Circuit saying which methods are proper when it comes to calculating royalties owed to patent holders - in cases where parties don't agree - determines what is reasonable when it comes to SEPs bound by FRAND terms. Relevant here, even if the Federal Circuit hadn't applied its entire market value rule in cases that specifically dealt with FRAND-bound patents, that rule would still dictate what was proper in such cases. The general rule would still apply.

    If anything, it would be going in the other direction where an argument could have been made that the rule didn't apply. If the court had only ever ruled that it applied in cases involving FRAND-bound patents, then it might be argued whether it also applies in cases which didn't involve FRAND-bound patents. Indeed, that was one of the issues addressed in CSIRO v Cisco (Federal Circuit, 2015). Do the rules from Ericsson apply even if the patents in question don't carry FRAND obligations?

    But, to be clear: Ericsson v D-Link (2014) - a case which addressed a number of legal issues - involved patents that carried FRAND obligations. And in that case the Federal Circuit reiterated the entire market value rule which I've been referring to. It said, among other things (citations omitted):

    Thus, where the entire value of a machine as a marketable article is “properly and legally attributable to the patented feature,” the damages owed to the patentee may be calculated by reference to that value. Where it is not, however, courts must insist on a more realistic starting point for the royalty calculations by juries—often, the smallest salable unit and, at times, even less.
    The Federal Circuit has now reiterated this rule in a number of cases and contexts. And it follows from more general Supreme Court precedent that's more than a century old. That is the state of the law as it exists now. It could, of course, change in the future. 



    IMO following the decision in Unwired Planet v. Huawei (and confirmed on appeal) may make it a global licensing issue rather than one decided under US law. 
    "Just as implementers need protection, so too do the SEP owners. They are entitled to an appropriate reward for carrying out their research and development activities and for engaging with the standardisation process, and they must be able to prevent technology users from free-riding on their innovations. It is therefore important that implementers engage constructively in any FRAND negotiation and, where necessary, agree to submit to the outcome of an appropriate FRAND determination."

    "When a patent holder declares a patent essential to an ETSI telecommunications standard, it must make a commitment to license that patent on FRAND terms. In this judgment, the court has held that a FRAND licence will typically be worldwide and that the terms of that licence should not vary substantially depending on the size of the licensee/implementer...
     
    The court confirmed that both parties must take a FRAND approach; the SEP holder must not hold up (effectively refusing to license) and the licensee/implementer must not hold out (effectively refusing to take a licence).  However, this does not mean that all offers made during negotiations have to be FRAND. A FRAND approach still allows for ‘non-FRAND’ opening offers that leave room for negotiation upwards or downwards towards an eventual FRAND licence.  Offers above or below FRAND will only become problematic if they are extreme or take an intransigent approach which disrupts or prejudices the negotiation. Notably, in the judge’s view, it is not hold up for an SEP holder to seek to obtain through its licensing some of the value that results from the inclusion of its technology into a standard.

    As a result of a particular issue raised in this case, the court also dealt specifically with the ‘non-discrimination’ or ND part of FRAND.  He held that a licensee/implementer is not necessarily entitled to demand the same licence terms as another similarly positioned licensee (although it may be able to if the differences in licence terms distort competition between the two licensees).  Instead the ‘non-discrimination’ part of FRAND means that licensees should be offered FRAND royalties based upon the strength of the patent portfolio and not the size of the licensee."

    Qualcomm's patent portfolio would be considered strong would it not, and has demonstrated that it is willing to negotiate and therefor not "intransigent? Further, as far as the US is concerned, wouldn't the SEP licensing terms set out in the contractual agreement with ETSI that made the portfolio available to willing licensees on a FRAND basis carry extreme weight? Those royalities were agreed to be based on a percentage of the finished device cost, as was stated by contributors from LG to Motorola, to Huawei to Nokia, and ZTE to Qualcomm. While the specific percentage may be modified as part of negotiations between the parties I don't see how the companies who contractually agreed to the terms under which the IP would be contributed can now be found to be acting illegally if collecting royalties on the finished device. 

    I'm in no way saying I personally think it should remain unchanged, but it is what it is. Tossing out a standards contract before the end of its term will come with a new set of problems going forward. I think that's one reason among many why no US court has done so and won't. 

    So I don't see Qualcomm at this stage being globally barred from collecting royalties based on a completed device cost, particularly with the limitations (some new) they've agreed to in accordance with the standards organizations, ie tiered royalties and a reduced upper limit on the cost. Nor do I see any way that Apple can avoid paying royalties to Qualcomm on a per-device basis any time in the near future even if not using any Qualcomm hardware.  While highly unlikely in the US in at least three venues now, China, Germany, and the UK, unwillingness to pay royalties for the use of standards essential IP can be grounds for an injunction on sales of the offending product, an outcome Apple will avoid if push comes to shove. IMO IP holders being forced into negotiating different licensing rates and terms across various markets is also becoming less likely based on recent rulings. 


    A few things...

    (1) Again, no one is suggesting that Qualcomm collecting royalties based on end device cost is illegal. 
    That was exactly the claim made by several posters in this thread which is what led into the discussion you and I are having. :)
     ...and yes, you have several valid points regarding recent court decisions, but to date those have all been case-specific involving individual contracts unique to particular parties, not FRAND licensing in general, the Unwired/Huawei ruling being an exception IMO. Many end up being at opposition if delving into the details, the TCL v. Ericsson one you brought up and Unwired Planet v. Huawei that I referred to for example.

    Still as you say there's much yet to be ruled on before there's clarity in the matter of SEP licensing. There's not even a definitive "what is FRAND" yet. By the way there's an excellent interview over at IAM involving multiple IP attorneys with expertise in SEP commenting on some of the recent cases which you might have an interest in. https://www.iam-media.com/frandseps/sep-licensing-standard

    As always I thoroughly appreciate your posts on the subject. Thanks! Great having knowledgeably folks involved in discussions, particularly those like you who take time to try to understand the topics. 
    As for court decisions being case-specific... yeah, each decision is specific to the case it is deciding. But they establish (or follow) legal principles and rules which are applicable to other cases. And the rule which I've referred to - the entire market value rule (which represents an exception to a broader principle) - is one that applies generally. It isn't limited in applicability to the specific cases I've referred to.
    Things are not as many here seem to believe they are when it comes to smallest saleable unit arguments as the SEP royalty basis. Even very recent Federal Court rulings have reiterated that royalties based on the end-product price rather than a component can be perfectly proper and thus legal, and on grounds beyond those you brought up in the
     Power Integrations v Fairchild Semiconductor International case.

    "The Federal Circuit’s recent Exmark v. Briggs-Stratton decision further confirms that there is no categorical rule about selecting a royalty base when litigating a reasonable royalty in order to apportion value to the patented invention, but that “apportionment can be addressed in a variety of ways … [s]o long as [the patent owner] adequately and reliably apportions between the improved and conventional features” of the accused product.  Thus, in this case, the Federal Circuit ruled that the patent owner properly could use the entire lawn mower as the royalty base and was not limited to the innovative baffle component of the lawn mower as a royalty base.  This case continues the clarification made in the Federal Circuit’s CSIRO decision involving standard essential patents that recognized that “adaptability [in determining patent damages] is necessary because different cases present different facts” and rejected as “untenable” the argument that every damages model must start with the smallest salable patent practicing unit"
    https://www.essentialpatentblog.com/2018/01/federal-circuit-confirms-flexibility-determining-royalty-base-exmark-v-briggs-stratton/#more-7848
    I don't see where that ruling conflicts with what I've repeatedly said. And to be clear, it doesn't involve FRAND-bound SEPs, for which the realities of a hypothetical negotiation are quite different.

    As I've said repeatedly, royalties based on the entire market value of products aren't illegal. Parties can negotiate and agree to use such bases. Further, settlements or negations which use such bases can be introduced at trial in order to support damage arguments. That's an evidentiary issue. I've never suggested they can't and I'm pretty sure somewhere I've specifically addressed that issue in relation to the Ericsson v D-Link decision.

    In this case, according to the Federal Circuit's decision, the patent in question related to the lawn mower as a whole. The entire product was infringing, not a single component of it. (That's due in part to how the patent is constructed.) That's a different situation from what we've been dealing with, and it's a relevant difference. When it comes to the Apple - Qualcomm situation (and to other cases which I've cited), we're talking about patents which relate to single components of multi-component products.

    If a patent relates to the entire product (as in the present case), then (when it comes to damage determinations) there still has to be an apportionment between the value of the patented element and the value of non-patented elements. But that can be done using the entire market value of the product as a base.

    From Exmark v Briggs & Stratton (citations omitted):

    Using the accused lawn mower sales as the royalty base is particularly appropriate in this case because the asserted claim is, in fact, directed to the lawn mower as a whole. The preamble of claim 1 recites a “multiblade lawn mower.” It is not the baffle that infringes the claim, but rather the entire accused mower. Thus, claim 1 covers the infringing product as whole, not a single component of a multi-component product. There is no unpatented or non-infringing feature of the product. 

  • Reply 38 of 39
    gatorguygatorguy Posts: 24,176member
    @carnegie , several of Qualcomm's wireless patents including some that are SEP-pledged read on the function of the entire mobile device and not solely its individual components like a wifi chip. How would that differ from the Briggs case? 

    Can you remove all of Qualcomm's inventions from an iPhone and still have an iPhone? 
    edited December 2018
  • Reply 39 of 39
    carnegiecarnegie Posts: 1,077member
    gatorguy said:
    @carnegie , several of Qualcomm's wireless patents including some that are SEP-pledged read on the entire mobile device and not solely its individual components like a wifi chip. How would that differ from the Briggs case? 
    It would likely, if Qualcomm wanted to make certain arguments based on that, require factual determinations to be made. Since a court would be looking at a group of patents, and trying to figure out damages for all of them, I suspect it would in essence boil the question down to this: Do these patents really relate to the product as a whole or, for the most part, to particular components of it? I'm not saying that's for sure what a court would do, but it would be my guess if I had to make one.
Sign In or Register to comment.