Apple tells its side of the story in Imagination Tech spat

Apple's dispute with Qualcomm is making plenty of headlines this week, but Qualcomm isn't the only supplier with whom Apple is having a bit of a spat. Back in April, Imagination Technologies announced that Apple had decided not to use its PowerVR GPU technology in upcoming A-series SoCs. Today, Apple gave a statement to Bloomberg that not only challenges Imagination's account of the timeline of events between the two companies, but also provides information that has the potential to be legally damaging to the British firm.

When it first revealed that its relationship with Apple would be coming to an end, Imagination suggested it was caught off-guard by the decision. Apple had been using the PowerVR graphics IP in various devices since the release of the first iPhone, and the majority of Imagination's revenue came from its contract with Cupertino. Since Apple's announcement, Imagination's stock price has fallen by more than half of its prior value this year.

In its statement today, Apple tells a different story. The company says it informed Imagination that it was winding down their relationship as early as 2015. Furthermore, Cupertino says that in 2016, it exercised a clause in its contract with Imagination that allowed it to pay a lower licensing rate by using less of Imagination's intellectual property. For its part, Apple said that it "valued" its relationship with Imagination and wanted to ensure the IP licensing firm would have enough time to adjust its plans in light of the impending end of the companies' deal.

Whichever side you believe in this matter, the situation is "almost certainly" going to be investigated by financial regulators, according to a barrister in touch with Bloomberg. That legal expert said Imagination has a responsibility to promptly inform its shareholders about events that could deeply affect its share price under British law, and if Imagination knew about Apple's plans in 2015, it may not have met that standard of disclosure. The situation is also likely to be complicated further by Imagination itself, who has publicly expressed doubt that Apple can design its own GPU architecture without infringing on its intellectual property. Whatever sad chapter comes next in this drama is almost certainly going to play out in a courtroom.

Comments closed
    • tanker27
    • 3 years ago

    I always wondered why Apple didnt outright buy Imagination. And Imagination playing like a spoiled child by claiming that ‘if’ Apple went on to design its own GPU architecture that it would most certainly infringe on ‘their’ IP.

    Who says Apple is going at it alone? Maybe they want to team up with someone else and use their GPU because, just maybe, Imagination’s GPU is limited or stagnant in what Apple wants to do in the future. I believe that this is the true case of why Apple decided to end the relationship.

      • tipoo
      • 3 years ago

      Always possible that they’ll licence some AMD or Nvidia patents, as Intel does for its GPUs.

        • tanker27
        • 3 years ago

        And that’s where I see this going.

        • NoOne ButMe
        • 3 years ago

        AMD.
        Doubt Apple is going to go to Nvidia.

        Despite NVidia currently offering products which without specific binning is equal in pref/watt to AMD’s offerings.

        Based on Vega’s power usage on the FE, it seems to me that Apple really doesn’t want to use Nvidia. Or they’re just gonna downclock like mad to fit into an iMac.

          • tanker27
          • 3 years ago

          I wouldn’t bet or even guess right now. There are pros and cons to each and quite honestly it can go either way.

          • tipoo
          • 3 years ago

          Nvidia is playing gatekeeper to the market to Intel right now, it may not be the same situation as Apples GPU selections. It’s about patents. There was some rumor that Intel would switch to licencing AMD patents, but I don’t think anything came of it.

    • NoOne ButMe
    • 3 years ago

    Hm. I would think Apple is more likely to be telling more of the truth here.

    I expect both to obfuscate and hide as much as possible, but I view Apple as having fewer reasons to hide.

    As was mentioned before, Apple has a GPU team, A full GPU team, excluding anyone else they brought on after they hired the initial team.
    This team likely has finished their design, the finishing is probably when Apple started to hint Imagination. And Imagination’s release probably came after Apple fully validated it and okay’d production

    I don’t see what about Imagination besides potentially patents which would interest Apple. Maybe Apple would want MIPs for the hell of it?

    I think in that case, Apple would just design their own ISA instead, I would think.

    • ronch
    • 3 years ago

    I would love to have a graphics card powered by Imagination graphics. Come on, IT! You can do it!

      • K-L-Waster
      • 3 years ago

      Ok, suppose that one was available in the fall, and you had a choice between an IT card and Vega? Which would you get?

        • ultima_trev
        • 3 years ago

        If they made a Power VR Kyro III… Kryo III all the way. Vega is too power hungry and throttles too much.

        I’m sure Vega will still have a considerable edge in gaming performance but likely not proportionate to the power draw increase.

      • srg86
      • 3 years ago

      As a Linux user, I won’t be shedding any tears for Imagination if they go under. Just hope their engineers find new jobs quickly.

      • alrey
      • 3 years ago

      we already have PowerVR for PC back in the days of 3dfx.

        • ronch
        • 3 years ago

        Yes, but I’m talking about a comeback to the PC graphics space today.

    • Delta9
    • 3 years ago

    Kind of surprised AMD or Intel isn’t eyeballing these guys. AMD sold their mobile line a few years ago and I would think there is some decent power saving tech and rendering techniques in their portfolio that would sit well in their GPU designs. Intel seems like it would be in a position to buy as well. They always need graphics IP (considering they used Imagination graphics in their atom chips) and their own IGPs are used widely by OEM in desktops and notebooks where power savings is selling point. I am sure Intel needs every little bit with AMD having a competitive CPU to pair with their mostly superior IGP finally. That and I would think someone like a Samsung, Qualcom, Mediatek, or a Chinese firm that are working toward or already making GPUs would be interested in acquiring the architecture and patents behind some of the fastest mobile GPUs on the market over the past 10 years.

      • NoOne ButMe
      • 3 years ago

      Intel has given up on mobile.
      AMD doesn’t want to attack mobile.

      Market is pretty much:
      Self made SOCs
      Ultra-low-margin SOCs
      Qualcomm SOCs

      Even Qualcomm makes most of their profit from licensing.

        • NoOne ButMe
        • 3 years ago

        Plus, unlike Nvidia, AMD doesn’t have the money to waste on a worthless mobile SoC development which eventually won a game console.

        Assuming the repost of Semiaccurate story are true, (as original story true) than Nvidia’s lifelong profit on Tegra is likely still in parentheses.

        I read Nvidia’s fiscal report latest as Switch revenue being grouped in with gaming, so I believe Semi story probably true, provided accurate leak/repost.

    • ronch
    • 3 years ago

    If I were Imagination I would at least try to enter the desktop graphics IP instead of acting like a guy who wants to jump off a bridge because his girl dumped him.

      • DavidC1
      • 3 years ago

      That is quite harsh to say.

      Also the comparisons aren’t valid. This isn’t a romantic relationship, its business. The situation with PC is also much different than when Kyro was available. GPUs have large sized dies and AMD and Nvidia aren’t tiny companies that could suddenly be usurped by what’s practically a startup. They have deep relationships with software and game developers, and they have substantial funding just to develop stable drivers. It just isn’t reality anymore.

      GPU driver and software development on the PC is unlike anything else. Networking cards, motherboards, monitors, keyboards, mice, those devices can get away with 1-2 driver updates in addition to initial release drivers. Yea, it sucks, but that’s what companies making peripheral devices do.*

      *You’ll realize Intel is actually pretty consistent this way. Their driver support actually excels against other companies making drivers for peripheral devices. But compared to AMD/Nvidia, especially the latter they are behind just because the GPU development is so fast paced and so demanding.

        • ronch
        • 3 years ago

        It may be a bit harsh, but you don’t see companies hanging a For Sale sign on their window because their largest customer stopped buying. It’s almost as though they just wanna give up right away.

          • NoOne ButMe
          • 3 years ago

          Look at Imaginations revenue per source, and profits (losses).

          Losing Apple will cost then somewhere in 40-60% of revenue, and unlikely to change their costs much.

            • ronch
            • 3 years ago

            Yes we know that, but compare that to AMD’s Never [s<]Settle[/s<] Quit mentality.

            • NoOne ButMe
            • 3 years ago

            AMD is one of the two (well, 3) companies that can really compete in the HUGE x86 space.

            Plans were in place to spin off part of the company if Zen was not good enough.

            Zen may not be good enough still, but I think it will be.

            • DavidC1
            • 3 years ago

            AMD is different.

            When AMD lost marketshare it happened gradually. That allows the company to deal with it. With IMG they’ll lose it almost overnight.

            Probably based on nostalgia people want IMG returning to PC graphics, but its not realistic. The big thing they had going on with Kyro was TBDR. Competition was barely into occlusion reduction techniques, now both AMD/Nvidia graphics are quite advanced there. They are also using some form of tile-based rendering, probably not deferred although its questionable how much of a benefit deferred part gives nowadays. Considering the complexity in rendering its said to bring, likely not worth it.

            Graphics is far more challenging than it was back then, not only in the hardware itself but the development of software. So it makes no sense to pursue opportunities somewhere else in mobile but they should go headfirst against much more formidable AMD/Nvidia in the PC space? Yea, right.

            There’s a reason Kyro died. Unlikely it’ll come back.

            • NoOne ButMe
            • 3 years ago

            ATI/AMD also fluctuated, before it’s drastic lost in marketshare.

            If only the initial version of Kaveri had been brought to market, wouldn’t have changed Nvidia’s dominance in dGPUs, but it truly could have killed the low end market.

            • K-L-Waster
            • 3 years ago

            AMD didn’t have one single customer who bought up 60+% of their chips and who had defined the specifications for said chips.

            The Imagination situation would be a bit like if AMD had been reduced to only selling their custom SOCs for the XBox and the Playstation, then MS and Sony suddenly decided to switch suppliers.

    • deruberhanyok
    • 3 years ago

    Anyone running a small business really needs to look at Imagination’s situation as an excellent example of why you should never have so much of your revenue depend on a single customer.

      • ronch
      • 3 years ago

      Given how they’re not 100% dependent on Apple shows they knew not to rely on them too much. Unfortunately many other GPU IP folks like Qually and Mali (which ARM probably tried to bundle with their CPU cores aggressively a-la Intel) pushed their graphics just as hard, if not harder. It’s not a walk in the park to push anything these days: competition is cutthroat.

      • ET3D
      • 3 years ago

      The question is, when does that make for good business advice? When should a small business say ‘that’s a big client, we shouldn’t work with it’ or ‘they upped the orders, it’s better to cut ties now’?

        • blastdoor
        • 3 years ago

        Exactly. It’s silly to say no to more revenue and profit.

        Instead, you need to manage the risk. Maintain a healthy pile of cash so that you can weather any storms.

          • Chz
          • 3 years ago

          It’s a classic problem that any company that’s ever supplied Wal-Mart is familiar with.

          Wal-Mart comes by and says “That’s a nice product you have there, and we’d like to sell it. The problem is that you charge $x for it. If we buy 50% of your annual production, we want to pay 0.85*$x.”

          So you think, at 150% of current sales it’s well worth it to give WM a price cut.
          Next year, Wally and Co. come back and ask for twice as many. But at 0.8*$x.
          Well, you’ll have to invest in expanding production, but business is booming and you can still squeak a profit. Sure!

          Next year, 50% more units and 0.75*$x. Tight, but doable.
          Next year, 0.7*$x. At this point, you’re not actually making any money. But Wal-Mart is 90% of your revenue and it would be impossible to pay back the loans on the new production facilities without them.

          Next year, you’re sunk.

        • deruberhanyok
        • 3 years ago

        The lesson is not to forego the revenue in the first place. The lesson is to take that revenue and invest it in slowly, intelligently growing your company, expanding your customer base and putting your company in an overall better position than they were before you had all of this revenue.

        Put another way: Don’t put all your eggs in one basket. Don’t rest on your laurels. Diversify your savings. Etc.

        I’ve seen a lot of small companies do this, they get that one big contract or land a huge customer, everyone is real happy, and then the next year rolls around and they’re going, well, they didn’t need to buy anything from us this year so our revenue is down by 30% year over year and we’re going to have to downsize…

      • Horshu
      • 3 years ago

      Or partner with Apple, ala GT Advanced (can’t really say that one was Apple’s fault)

    • Sahrin
    • 3 years ago

    I don’t understand why Apple didn’t simply buy IT before all this went down. Apple’s SoC GPU’s are almost always the most performant of their generation – meaning the Imagination IP is good stuff. You could but it all, or…reinvent the wheel? At the very least for the IP portfolio.

      • NoOne ButMe
      • 3 years ago

      Because Imagination Technologies current GPUs weren’t what Apple wanted.

      Apple would require Imagination Technologies team to rebuild GPU probably. And already they hired a team.

        • the
        • 3 years ago

        From previous rumors, [url=https://www.macrumors.com/2013/05/29/apple-hires-at-least-a-dozen-former-amd-employees-for-an-orlando-gpu-team-still-hiring-more/<]they hired a team several years ago.[/url<] Even Imaginations own COO left them for Apple last year.

        • green
        • 3 years ago

        plus if there was anything they wanted from them, then when you’re the company’s majority source of revenue it’s much easier to bankrupt the company and pick up the ip and any talent in a fire sale than to try buy a majority stake (as the market will react to your takeover bid and force you to pay a premium on shares)

          • brucethemoose
          • 3 years ago

          There’s no guarantee that there will be a a fire sale though. Someone could easily beat Apple to the punch when they go under, whereas buying them out would be a safer bet.

            • Zizy
            • 3 years ago

            I guess Apple got many folks from IT already, full company would be just a burden.
            Sure, IP would be nice to have, but when did Apple care for any other IP than theirs?

    • the
    • 3 years ago

    While a court battle is seeimgingly destined to happen, a company can’t survive solely as a legal defendant for long. I’m wondering what they’re doing in their labs as a last ditch effort survive. A hail mary pass with their Caustic ray tracing accelerator? Attempt to re-enter the PC graphics market place?

      • crystall
      • 3 years ago

      ImgTec’s recent IP has been doing fairly well outside of Apple’s so they can probably stick around for a while.

        • NoOne ButMe
        • 3 years ago

        maybe, but they have lots of fixed costs, in part due to them being IP based, and without Apple’s business, well.

        I believe Apple contributes half of their revenue, give or take. And losing that will have a negligible impact on their spending.

        I do think we will see them spin off PowerVR into a seperate company, or the rest into a seperate company.

        That part seems dragged down by the rest. But that might just be due to Apple…

    • windwalker
    • 3 years ago

    Yeah, I wonder which company had the incentive to lie: the one which lost its largest customer or the one who could have bought its supplier with what is for them loose change.

      • HERETIC
      • 3 years ago

      Excellent point-and I think your spot on.

      When your dealing with apple thro you have to remember-they live inside they’re own
      RDF and “honesty” is a concept totally foreign to them……………………

        • imnu
        • 3 years ago

        Indeed, we are talking about the firm that told the courts that

        “Before the iphone, cell phones were utilitarian devices that had numeric keypads and small passive displays with did not allow for touch control”

          • windwalker
          • 3 years ago

          That statement is true.

        • windwalker
        • 3 years ago

        You know you’re brainwashed when your conclusions change radically based on names of parties involved being available or redacted.

          • Beahmont
          • 3 years ago

          Or you understand that different things are different things and that the parties previous histories can greatly change the subjective comparison being implied in most arguments?

      • brucek2
      • 3 years ago

      There’s likely plenty of grey area here. There could have been dozens of times over the years that someone from Apple hinted it could work with other vendors if it didn’t get what it wanted. Clearly in many of those cases IT did get end up getting them what they wanted and the relationship continued. Finally there came a time when they didn’t.

      Apple could look back to any of those discussions and say “we gave notice.” IT could look back to the same discussion and say “we heard a negotiation ploy like we have many times before.” Neither would be lying.

      Of course I don’t know, maybe there really was a bright line formal notification unlike anything that came before it and that Apple stuck to exactly as delivered, but it’s certainly not out of the question that there wasn’t.

        • windwalker
        • 3 years ago

        Fortunately, courts can subpoena all relevant communication between the parties.

      • Beahmont
      • 3 years ago

      What a logical fallacy. Why would Apple pay when it can just steal the tech and kill Imagination Technologies with legal fees? Apple’s own legal fees under this scenario are still going to be smaller than the cost to buy IT, and even if Apple lost it would come years too late for IT to do anything about it.

      You comparison is a lovely example of the Apple RDF.

    • tay
    • 3 years ago

    It’s publicly not publically.

    This unwinding of the relationship is sad to see, but tellingly no one seems to be in a rush to pick up the powerVr IP

      • adisor19
      • 3 years ago

      Which is a shame really.. I think i’m not the only one hoping they would come out with a surprise Kyro competitor to Nvidia and ATI.

      Adi

        • rxc6
        • 3 years ago

        [quote=”adisor19″<]I think i’m not the only one hoping they would come out with a surprise Kyro competitor to Nvidia and ATI.[/quote<] Hello time traveler! I hope you enjoy 2017

          • meerkt
          • 3 years ago

          adisor19 may be misguided, but there’s no need to be condescending.
          Myself, I’m placing my bets on Glaze3D.

          • Srsly_Bro
          • 3 years ago

          Some people still live in the past and can’t accept the AMD buyout and absorption of ATI in 2006. It’s only been near 11 years so don’t rush him.

            • adisor19
            • 3 years ago

            For the Canadian in me, it will always be ATi.

            Adi

            • meerkt
            • 3 years ago

            Yeah, it was somewhat sad.
            At least they weren’t bought by Nvidia.

        • tipoo
        • 3 years ago

        I’d love some new blood in the desktop GPU wars.

        I’m not sure Imagination is really big enough to gain a foothold though, drivers and game support will be an uphill battle.

        Maybe instead, AMD or Nvidia could buy them out. Or even Intel to make better GPUs.

          • DavidC1
          • 3 years ago

          “Or even Intel to make better GPUs.”

          It’s actually quite easy to make fun of them in the graphics space because they purposely sell it for low end customers. But they are quite good in the segments they are at. They don’t have serious graphical and compatibility issues with latest PC titles anymore, and they do it at very low power.

          When there were more makers Intel was #3 behind AMDTI/Nvidia. While aside from AMD/Nvidia no-one else was known for 3D performance, the one that had passable general-usage 2D capability happened to be Intel. SiS and Via were just not good there.

          You are underestimating how much effort GPU development takes. It’s a much different world than when Kyro was in the streets. I’d bet even Intel looks good then.

            • tipoo
            • 3 years ago

            I’m not making fun of them, I’ve always followed their GPU developments with interest post HD series. Seems they have a pretty good architecture, if it was scaled up I wonder how it would do as a dedicated GPU.

            What Imagination can provide though is patents, they could perhaps more directly address certain problems than using workarounds that avoid all the older GPU makers patents.

            (I also don’t think they’re particularly efficient at peak load, iirc the Iris Pro Macbook Pros could use more power than the dGPU ones for less performance, they were more efficient at idle though)

            • NoOne ButMe
            • 3 years ago

            If they built their GPU on TSMC, Samsung or GloFo, or build a process themselves which wasn’t aimed at mobile or high end CPUs, they could get a pretty good product, I think.

            Intel’s process is just to CPU-focused to be really good for a dedicated GPU at the moment, I believe.

            • Zizy
            • 3 years ago

            They aren’t that efficient, when you look at comparable AMD products you see perf/W of AMD APUs was actually better in games even though their CPU side was much worse. Haven’t checked in the last 2 years though, could be Intel caught up further.

            But yeah, they are in the same efficiency ballpark at least, so they probably could make the GPU bigger and compete with desktop stuff… maybe, perhaps, who knows. It isn’t that easy to scale up and down and until they do it, we can’t know for sure if they can or can’t.

        • Zizy
        • 3 years ago

        The problem is that everything below 100W is barely worth making because of iGPU, and they don’t have anything that even appears to be high performance. Sure, they could go after some pro markets supporting tons of screens, but this would mean abandoning their own expertise and doing something entirely new for them.

          • NoOne ButMe
          • 3 years ago

          Nothing is stopping Intel from making a massive GPU.

          After all, they have the cheapest process and the best yields. If they ever have low fab utilization, what’s stopping them?

          Certainly a better use of money than their mobile jokes of products, and probably worth overpaying slightly less for their massive buyouts. I understand overpaying for stockholder/investor value. But Intel’s payments seem excessive to me.

            • Zizy
            • 3 years ago

            I wasn’t talking about Intel. They = Imagination Technologies.
            IT have a reasonably good low power GPU. Nothing says it scales nicely to desktop/laptop performance. And if it doesn’t meet at least about 460/1050 level of performance you might as well not bother making it, iGPUs take care of that market.

            As for Intel, their Phi started as a GPU and is now trying to find second life. Even if they have the best fabs etc, it doesn’t matter if they don’t have best GPU guys.

      • CuttinHobo
      • 3 years ago

      And not pubicly, either?

      • ImSpartacus
      • 3 years ago

      I make that damn mistake so often…

      I can forgive that typo.

      • ronch
      • 3 years ago

      Purportably.

Pin It on Pinterest

Share This