This is the very reason why I had loaded up on significant amounts of Qualcomm stock earlier this year. Apple pretty much had no choice but to settle a deal with them or get screwed on 5G.
I have no doubt that Apple will likely attempt to sever their dependency on Qualcomm when they make eventually make their own modems (they've made incredible strides with their SoC's under Johny Srouji), but it will still be at least a few years before that will even be a reasonable option.
The complexity of engineering that goes into a high-performance modem is nothing to sneeze at, especially when Qualcomm owns a significant portion of 5G-related patents. Even Intel threw in the towel upon the realization that the capital expenditure wouldn't be worth it, and it would be impossible to compete with Qualcomm profitably (Intel stock actually jumped up accordingly). Qualcomm will continue to collect licensing/royalties from other manufacturers like Huawei as well, and is in an excellent position frankly.
> Even Intel threw in the towel upon the realization that the capital expenditure wouldn't be worth it
Intel doesn't seem to me like a company that is able to embark on a new technology, innovate and surpass the competition. Take a look at dGPUs, Intel has been getting smashed about by Nvidia for decades and there's a fantastically long history of Intel slowing buying up and killing hundreds of small GPU companies. If I were to pick someone to move into the Modem space and succeed it wouldn't have been Intel.
Intel's integrated GPUs have different goals than NVIDIA's dedicated GPUs. Power efficiency and video decoding and encoding is more important to Intel than raw gaming performance.
> The complexity of engineering that goes into a high-performance modem is nothing to sneeze at
For Apple to make a 5G modem that's competitive on power and area with Qualcomm's will be a significant feat. But Apple's got some really good design teams and I'm sure they'll figure it out eventually.
Apple has an extremely competent team of chip designer, decades of experience in manufacturing and plenty of money to snatch top talents. If they want to be serious about modem design and it seems they want to, they will most likely succeed.
I'm not sure I follow you here. You seemed to be implying that after witnessing Blackberry failure to properly make their own modem you found the idea of Apple doing it laughable.
I was pointing to you that :
1. Contrary to Blackberry when they started trying Apple has been designing chips (CPU) for some time and has knowledge internally about how to conduct this kind of projects and how to build a team for it ;
2. Apple already understands and masters the manufacturing part of chip making ;
3. Apple is so rich and has such a brand that they can most likely buy the people mastering the knowledge they still lack (anything modem specific) from their competitors.
The idea that Blackberry was doing before everyone else seems to only make my points stronger. Apple will have less R&D and engineers training to do.
I'm not arguing it, I'm just saying it's a fool's errand.
If you build your own chip, you have to build your own mobile stack, and then maintain it and enhance it with the newer technologies, so it becomes a mountain of technical debt while your competitors can take what Qualcomm gives them and roll out a phone.
This assumes that Apple will maintain it's location as the top of the phone line and won't get eaten by someone else, if they slip, that mobile stack becomes a liability.
None of these advantages invent you a novel radio technology that is competitive with 5G yet unencumbered by existing patents. Designing a modem to spec is the easy part.
Slight modification to your narrative here (mostly all right) Intel through it in once Apple signaled there was a minuscule likelihood they’d go with one of their 5G modems anytime soon
It wasn’t entirely impossible to reach this conclusion through sensible speculation alone.
You don’t need to be Apple to know they only had three options for modems really:
1. Keep buying from Qualcomm.
2. Try and prop up Intel’s modem efforts in hope it becomes viable competitor to Qualcomm to either increase leverage or just get Intel good enough that you don’t need Qualcomm anymore.
3. Build own modems.
They tried to make 2. work, and 3. as others have noted simply isn’t viable for several years at least presumably.
Given Intel’s pretty poor track record at entering into new markets and succeeding, it wasn’t a bad bet that 2. and 3. was doomed to failure forcing a retreat to 1.
We already knew from hardware tests that Intel’s 4G modems are not competitive with Qualcomm’s too, so not unreasonable to speculate the future 5G ones might not be as good either. Thanks to Intel’s press release the other day we can largely confirm too that their 5G modems weren’t going to be competitive.
This article suggests there was also option 4, which is not something I’d have as easily speculated:
4. Try to engineer legal/market conditions that will negatively affect Qualcomm’s business to drive modem prices down.
It is this option that seems to have failed this week, if this article’s analysis of Apple’s strategy is correct.
This article is conflating two separate things: Apple said it should pay less for Qualcomm’s PATENTS. It always liked and never really said bad things about Qualcomm’s MODEMS. Patents != modems.
Also, the article quotes “Adam Mossoff, a law professor at George Mason University” without disclosing that Qualcomm funds GMU exactly so it will get favorable press like this.[1]
The FTC would have you believe that Qualcomm leveraged its patent portfolio in order to unload low quality chips. If that were true, then it would seem to be harming the marketplace. But in reality the chips are market leaders because they deliver superior performance/value, not because of patent leverage.
Obtw it's very likely that Apple applied leverage to convince FTC to bring their suit.
> The FTC would have you believe that Qualcomm leveraged its patent portfolio in order to unload low quality chips.
Citation needed. If anything, it's the reverse: the FTC complained that Qualcomm leveraged its chips in order to get inflated royalties for a low quality patent portfolio. From the original complaint:
> a. Qualcomm withholds its baseband processors unless a customer accepts a license to standard-essential patents on terms preferred by Qualcomm, including elevated royalties that the customer must pay when using competitors’ processors (“no license-no chips”).
(Edit: The "elevated royalties" part does imply Qualcomm was also trying to prevent its customers from purchasing its competitors' chips, but the way it allegedly negotiated the agreements with those royalties was by threatening to withhold its own chips. If Qualcomm's chips were low quality, a customer wouldn't have much to fear from such a threat. That is very much not the allegation.)
Yeah I know you don't really understand how things work. The patents only tell about half the story in the modem design. the part that Qualcomm leaves out of the patents and they leave out of the standards that's the part they put into the modems. It's why Intel's modems are garbage.
One of the comments on the site proper mentions that in light of the actions of the larger tech companies of the past decade, the commentor feels like they were unfairly harsh in their judgement of Bill Gates in the 1990s.
That's such a dangerous conclusion that's used in politics from time to time. Just because X is worse than Y, it doesn't mean that Y wasn't also bad. Y is only better by comparison. It's as objectively bad as it ever was.
Strive for better instead of simply settling for not as bad.
How is it dangerous? How are you able to equate "I was unfairly harsh in my judgement of X" with "X wasn't bad"? That doesn't follow, it's simply a comparison.
Gates and MS anti-trust violations are now standard operating procedure for today's tech giants, the behavior of MS in the 90s looks incredibly harmless when compared to the wildly irresponsible socially dangerous behavior of tech giants today. We seemed to have just collectively forgotten about anti-trust and decided that the natural monopoly of tech services will be the new status quo. But anti-trust is nothing compared to the perils and social consequences of the 'move fast and break things' approach of technology and software today. The Boeing 737max is one recent example, but we're just starting to unravel the social and cultural costs of allowing companies with unprecedented power to shape human behavior and public opinions, dictate what is and is acceptable speech for their platform in the name of quarterly growth. Seems like a reasonable comment and insight to me.
You're right that the irresponsible behavior of tech giants today makes what Microsoft did in the 90s look harmless. But only by comparison. It's dangerous because in this case hindsight is not 20/20. What Microsoft was doing in the 90s wasn't harmless. It was, at the very least, a step on the path that led to the very behavior we're condemning from others now. And besides that, their behavior was objectively harmful.
It's strange that ms got busted partly for their handling of internet explorer, but at least on windows there was never any restrictions on what 3rd party software was allowed and no obligation to pay Microsoft a 30% tax on sales. Meanwhile Apple has 50 pages of App Store Review guidelines, routinely rejects apps for political reasons. And collects taxes on everything.
In the 90's Windows had > 90% marketshare of the desktop market and Internet Explorer had > 90% browser marketshare. Globally, iOS is < 25% marketshare. So while Apple holds an app store monopoly on their own phone, you have a comparable alternatives available. Microsoft lost because, "The Court has already found, based on the evidence in this record, that there are currently no products - and that there are not likely to be any in the near future - that a significant percentage of computer users worldwide could substitute for Intel-compatible PC operating systems without incurring substantial costs." In retrospect, that looks questionable, but at the time Windows dominance looked pretty unstoppable.
Microsoft suits were not global. There is no global government. In the USA Apple is 50% +/- 5% in Mobile. Samsung is #2 at 24%, Lg 6%, Motorola 4%
On top of which Microsoft suit specifically calls out "Intel compatible PC operating systems"
From the ruling
> The District Court determined that Microsoft had maintained a monopoly in the market for Intel compatible PC operating systems in violation of 2; attempted to gain a monopoly in the market for internet browsers in violation of 2; and illegally tied two purportedly separate products, Windows and Internet Explorer ("IE"), in violation of 1.
In other words, Apple's market on Power PC machines at the time was not even taken into consideration.
Using the same logic we Apple has 100% market share on A-series based processors. ARM is not hardware it's a spec that any company can build hardware around and Apple is building their own.
Apple dictate the browser engine to 1.3billion devices. They use that monopoly to force browser standards since by disallowing browser engine competition on iOS they are able to lord those 1.3billion users over the any standard they want/don't want since other browsers are iOS are not allowed to provide their own engines and therefore can't support newer standards and so Apple prevents and real competition in browsers on iOS.
Similar issues are coming up with Apple being the only store and only way to get software on iOS. Those issues are in the courts now in both the USA and EU.
On top of which Microsoft suit specifically calls out "Intel compatible PC operating systems"
They are talking about Companies like Dell, Gateway, Compaq, HP, who made intel hardware. What choices did those companies have for an OS for their computers? Windows effectively had a 100% monopoly. They aren't talking about consumer alternatives, where you could by a Mac.
55% isn't a monopoly, Android is a very good substitute, so not really comparable to the 90s where you needed a desktop and Office to get work done and Windows controlled 90-95% of the market.
Maybe because some people on HN don't like to be reminded that their bias against Apple doesn't play out in what the law actually states about monopolies, and the factual market across the globe.
poor phrasing on my part and I think we're in agreement about this, of course companies should have control what goes on their platform, they're not public services - the issue today is they have the power but none of the consequences from potential fallout. Should we really be placing that power in the hands of for profit entities? Regulation is it's own bag of worms, but at least the government is in theory meant to be acting in the interest of the populace vs. shareholders - a key difference especially in dealing with influence/reach/impact of the likes of twitter, facebook, reddit etc
These statements aren't exactly mutually exclusive. Most folks in the industry generally accept QCOM is the best, but also that they are no good. They're borderline patent trolls. While they do some good, the market would've advanced much faster without them.
By market, who are you talking about? The Europeans? Seriously? The Asians? With TD-OFDMA? Seriously? Intel? Intel has never been a successful network company. they bought every goddamn Network product (through acquisitions) they ever sold in the marketplace.
When one company dominates a technology, it reduces incentives for other companies to join or be founded to address that space. If Qualcomm did not exist, there would be a market vacuum of that size, and it is impossible to predict how that vacuum would end up being filled decades later.
The hope is that there would have been an entrant with similar technological prowess, but more humane business practices.
(I have personally worked with Qualcomm in the past on both technical and business fronts. Their technology can be very good, but their business side can be straight up offensive in their anti-competitive and predatory demands. I would love to be deposed on this topic.)
I was actually quite disappointed that they settled at the last minute, as we no longer have access to juicy legal discovery that would have revealed Apple's rotten legal strategies and bad-faith practices. Now it's slowly becoming apparent that Apple fabricated evidences and tricked Qualcomm into legal troubles with regulators (eg, rebates for exclusivity) -- which is why QCOM stopped paying Apple rebates.
I don't think Apple is giving up its unscrupulous practices, however, and I'm pretty sure they are setting themselves up for the next battle years from now (Apple is now a direct licensee). In hindsight, QCOM should have taken lessons from Samsung's battle with Apple years back. There were so many red flags from the moment the FTC approved antitrust investigation of QCOM just days ahead of a new administration with 2 of the 5 commissioners missing, against the vehement dissent from the interim chair. I just have to wonder what's in their new licensing/business agreement -- if QCOM had to make concessions like it did before and it's another legal or regulatory minefield.
At least on paper, Apple's work on ARM chips is truly incredible. I recognize that it's not gonna happen but I would kill for a dev board with an A12X, and I doubt I'm alone. (There wouldn't be much of a point to one in practice, but I bet it would make a very formidable low power desktop.)
> I bet it would make a very formidable low power desktop
Perhaps if Apple really transitions to ARM on the Mac Line, that might be a reality in a couple years. I'm not sure they would want to replace x86 for the 'Pro' Desktop machines, but I don't think an ARM Mac Mini is out of the question.
ARM for Mac depends upon the Software ecosystem. Though much of enthusiast tools are ARM compatible, in the Linux/ARM scene; professional software are long way from it.
Unlike a dev board, Mac caters to professional audience. Apart from iOS apps, Several professional x86 software needs to be ported to ARM before ARM Mac becomes a reality.
Yes, Mac ARM chips are likely to feature some x86 emulation.
On a side note, I wonder about the patent dynamics regarding x86 & x86-64 emulation.
When Microsoft & Qualcomm announced snapdragon powered Windows 10, intel gave a 'not-so-subtle' threat to Qualcomm regarding patents for x86 emulation[1]. MS, still released the model anyways.
I guess, Apple has enough leverage to negotiate patents for x86(intel),x86_64(AMD) emulation.
> so they can put 48, 96 cores in a single machine.
> so just add more cores to it.
Oh my...thermals. I don't know where to begin. Scaling cores (even if they're efficient) poses many challenges - thermal, core-to-core bandwidth, interposer design issues and warpage, and most importantly, yield as the die size gets larger. Or you'd have to break up the core dice into many die and create fabric such as what AMD Epyc server chips are doing, etc...etc. A 96 core die with decent power envelope would be in the order of 300-500 watts.
There is no shortage of software to run on a sufficiently fast ARM processor. [0] The bigger problem is that (even without emulation) most ARM processors aren't powerful enough in comparison to x86 processors. Apples processors are the exception and they are only available with locked down operating systems.
> Apart from iOS apps, Several professional x86 software needs to be ported to ARM before ARM Mac becomes a reality.
While this is true, the iPad Pro is the key. If Apple can continue to get pro apps ported to the pro they will have many already moved to ARM by the time a full machine comes out. See many Adobe apps already out or in progress.
This is it. Apple follows a mobile methodology for their chip designs (tall: very powerful, large size cores) while Qualcomm, Samsung, etc follow an embedded methodology (wide: many small efficient cores). On top of that, Apple has some great CPU engineers who've worked on the 68k and POWER architectures which they've leveraged against their ARM design. They're also the only consumers of their CPUs which allows them to design them against their use cases and optimize for them exclusively.
While I'm a big fan of ARM in general, the only other chips really competing against Apple in it's performance + efficiency are nVidia's Denver cores and the ThunderX2; for similar reasons. Not that Qualcomm or Samsung's chips are bad by any means; just different.
I wish Qualcomm took all that legal revenue and converted it into a fab. They are basically limiting their upside now. Patent licensing for wireless standards has to be "fair". If the LTE/5g patents were open, they could charge the max possible amount for a fully integrated arm + modem core optimized down to the material science.
Apple has a LOT of servers. They just don’t sell them (well, they do). But they care a lot about it; they hired the Netty team to work on fundamental server software stacks (which we know because it’s open source). If ARM servers mean more power efficient data centers then I can see them doing it just for their sustainability objectives which is a major priority.
If an Apple-developed server SOC was put into wide use internally for sustainability reasons, wouldn’t Apple be morally obligated to open source such a design?
It seems like a possible conflict if it had tech otherwise proprietary to the A series.
Apple had given Intel modems a shot, and then users started complaining about performance differences. In some cases the Qualcomm modems would outperform the Intel ones by over 70%, especially in situations with low/weak signal. To maintain consistency and avoid outrage they tried to intentionally nerf the Qualcomm modem performance in these situations (since the Intel ones were lagging behind significantly).
Furthermore, the iPhone X was intentionally only made with two antennas, even though the Qualcomm X16 supported 4x4 MIMO and thus gigabit LTE because Intel's modem at the time didn't support that. Once Intel could do it, they released the iPhone Xs which is the same body but with the correct antenna support.
I worked for Qualcomm on 802.20. This was their first OFDMA system, complete in 2007 but polished for release at a disastrous time - 2009 - and all the patents were eventually fed into LTE. LTE, being a European standard, was 4 years late and 80% as good ...
The company in 2002 used a large compute farm to simulate the entire network stack (57 cells, hundreds of users) and to prototype changes to improve sector throughput. My group built the network emulator and tested TCP performance. They tested and simulated the crap out of their designs. Once the hardware was all built then they drive through the canyons of San Diego which is the second-worst wireless environment on Earth (Hong Kong #1 worst) to tune it so that it works well. Then they charge "too much."
The Europeans slap something together, producing networks very sensitive to site planning to work AT ALL, and then they charge a lot more for the network hardware changing data networks the way a young girl changes her shoe color every other day...
You have a choice when it comes to wireless network hardware and software. Qualcomm stuff is Rock solid. No other company does half of the design and field proving the Qualcomm does with their Network equipment.
As The European working on replacing firmware from another nation (with a little reverse engineering here and there) I can totally understand that attitude. When you make something really polished with best practices or somethng trully innovative, seeing what rushed underpaid/overmanaged programmers do to code evokes only disgust. Specific examples would be several hours of talking.
I thought it was 'Se Germans'. We make progress here.
In another of your comments it's "The Asians". I don't think it helps putting billions of people into a bucket and then label and bad mouth them. It doesn't explain anything about the world and helps no one.
"second-worst wireless environment on Earth"
Well I have splendid working 4G here in Europe. If that's the second worst wireless environment on Earth I'm fine with it.
Indeed it is utterly broken as up until March 2013 it was first to invent allowing the little guys/girls to get their fair share. Upload a demo video of their creation on YouTube and boom you were protected if there was no prior art.
Now it's first to file so if your an inventor inventing you better have the thousands of dollars to run to the patent office before some rich lazy tool steals your blood, sweat and tears.
Overall the tech companies helped/lobbied to get this terrible law passed!
Anything which is published before your filing date is prior art. Something sitting in your garage that only you and your drinking buddies know about or something your coworkers have been chatting about on a whiteboard at the office is not prior art.
This also means that if you want a patent you need to file a provisional application before publishing your work anywhere (e.g. the other commenter’s YouTube example). https://en.wikipedia.org/wiki/Provisional_application
Having a first-to-invent patent system makes patent disputes into complicated messy legal fights about when different people first encountered an idea. First-to-file leaves much less ambiguity.
Less ambiguity is not automatically an improvement. I'm not convinced that the lowered ambiguity of first to file is a net win.
Even with the availability of provisional applications, the 12 month limit on priority date, and the relative stringency required to avoid challenges to the priority date, doesn't seem to tip the scales significantly towards encouraging the creation of what we think of as inventions while simultaneously simplifying possible court proceedings, or lowering ambiguity as you put it.
I wonder what all of this is like from an engineer’s perspective. I imagine a lot of internal resources were aimed at the problem. There must have been a tremendous amount of pressure to ship something that was ultimately thwarted by the quality of Qualcomm’s product. I’m no hardware engineer, and I’m sure some of the particulars are private... but there is probably a remarkable story there.
I have a couple of phones, and the iPhone X has by far the worst Wifi connectivity. My Moto X4 and an older iPhone 6 Plus can both connect to WiFi through walls and other barriers. My iPhone X's loses its WiFi connectivity almost as soon as I leave my room. Even GPS accuracy is unacceptably horrid (making it useless for driving) on the X compared to my old 6 Plus.
It's saying something because my iPhone X is the most I've ever spent on a phone. With AppleCare NYC sales tax, the 256GB iPhone X cost me ~$1,450. Approximately what I paid for my MacBook.
I don't why the X's connectivity is so poor, buy my best guess is they changed the WiFi/network chip away from Qualcomm.
I find it odd that the article is implying that the only reason for Apple to negotiate for cheap patent licenses elsewhere is to artificially portray Qualcomm as overcharging, when surely it's in Apple's interest in general to negotiate favorable patent licensing deals.
Even if Apple's sole goal was to prove that they could license a much larger portfolio of patents for much less as a way of demonstrating that Qualcomm is overcharging, the fact that this was their goal doesn't make them wrong. "I'm going to prove you're overcharging by demonstrating that I can license other patents cheaply. Look, I licensed other patents cheaply!" is actually a pretty good argument that Qualcomm is overcharging.
The difference is the supporting documents found by Qualcomm during discovery. Internal Apple documents stated that Apple was actively trying to portray Qualcomm's patents as inferior by going to other providers.
Where does it say that? I've read the article twice now, and nowhere does it say that Apple internal memos tried to say Qualcomm's patents were somehow inferior. What it said was they intentionally licensed less-expensive patents to try and make Qualcomm's look more expensive, but that's the argument I already addressed.
That's exactly what the FTC did during their trial. The FTC lined up QCOM's competitors and had them testify that QCOM's patents were insignificant/inferior compared to their own patent portfolio. I'm fairly sure that this repertoire was in Apple lawyer's play too.
I can't seem to find an article talking about what you just described. Though I'm not sure what the FTC did even matters since we're talking about Apple, not the FTC.
No. the problem here wasn't that Apple was a tough negotiator, but that it also lied to regulators around the world that QCOM violated their FRAND obligation by overcharging.
The FTC obviously had no such proof that QCOM was "overcharging" or breached their FRAND obligation, so instead they used the euphemism like "elevated fee" to describe QCOM's licensing practice. In other regulatory investigation, QCOM wasn't even allowed full access to Apple's false testimony (eg, rebates for exclusivity or Apple's switch to Intel), much less representation to respond to its accusation.
Well it points out that they certainly aren't suffering from ability and don't need to try to skimp on the licensing. There are limits but it points out that trying to keep the price down wasn't necessary.
Now the true definition of 'overpriced' in this context is 'can they have some alternative that can do a better or cheaper'? Both will try to maximize their own profits and it is possible Apple made a penny wise pound foolish plan to try to cut costs.
I have no doubt that Apple will likely attempt to sever their dependency on Qualcomm when they make eventually make their own modems (they've made incredible strides with their SoC's under Johny Srouji), but it will still be at least a few years before that will even be a reasonable option.
The complexity of engineering that goes into a high-performance modem is nothing to sneeze at, especially when Qualcomm owns a significant portion of 5G-related patents. Even Intel threw in the towel upon the realization that the capital expenditure wouldn't be worth it, and it would be impossible to compete with Qualcomm profitably (Intel stock actually jumped up accordingly). Qualcomm will continue to collect licensing/royalties from other manufacturers like Huawei as well, and is in an excellent position frankly.