Hacker News Reader

Comment by gchadwick | Hacker News Reader

I have only skimmed the article but two things jump out:

> Recently the company got a fresh incentive to go all-in on silicon: revelations that microprocessors with components designed by Intel Corp., Arm Holdings Plc and Advanced Micro Devices, Inc. are vulnerable to hacking.

Apple isn't immune: https://support.apple.com/en-us/HT208394 (indeed seems iOS devices suffer from Meltdown where AMD devices do not).

> The result: a chip powerhouse that could one day threaten the dominance of Qualcomm Inc. and even, eventually, Intel.

Surely to threaten dominance Apple needs to start selling it's chips to other people? Whilst it's possible it would run counter to the way Apple likes to do things, namely get as much in house a possible to get maximum control and seamless vertical integration, they've never shown any interest in selling bits of their technology to others (well recently anyway).

View on HN

Apple may be a (large) minority purchaser of modem chips overall, but it's a hugely dominant chunk of the market for high end devices. If they do their own high end modems, it would gut the high end modem business for Qualcom. That could also drive up prices and jeopardise supply for other high end phone manufacturers. It might even make some high end features uneconomical for Qualcom to include.

Right now Apple is several years ahead of other high end phone manufacturers in CPUs. If they cripple Qualcomm's high end modem business by taking most of it away from them, they could extend the same lead to modems.

> Right now Apple is several years ahead of other high end phone manufacturers in CPUs.

I hear this statement a lot... how is this determined?

Twitter pundits commenting on Geekbench benchmarks, mostly.

Snark aside, Apple's CPUs are pretty(really pretty) far ahead of Qualcomm's in single-threaded performance.

Qualcom's chips tend to struggle to keep up with whatever Apple released two generations ago.

But it doesn't matter because if you're building an Android device you can't buy an A chip for it.

Given how highly multithreaded mobile OSes are, that is only advantageous for lazy coding.

Any proper iOS application with be doing lots of GCD dispatchs, for starters.

Likewise on Android, Google had to change the OS behavior to just kill apps that insist in misusing the UI thread for long running tasks.

Single threaded programming isn't 'lazy coding'

It is, in the age of ubiquitous multi-core hardware.

I don't remember when it was the last time any of my applications only had a single thread of execution on them, beyond shell and Python scripts.

Maybe around 2000.

And even then, the OS is juggling processes across all cores every few ms, so outside any benchmark winning game, there isn't much real world value in single thread performance.

Except for JavaScript performance: https://twitter.com/codinghorror/status/954493740600578049

What I see is the quality of a JIT compiler implementation, not the use of cores.

Are you serious? Multithreadjng is not Parallelism. It’s not about processing an image in the UI thread, it’s about making that processing faster on the background thread that picks up the work. Don’t get me wrong but lots of people still write iterators, for loops and single threaded functions. Parallel algorithms are hard

Uh nope, single threaded performance is the key parameter. It’s what made intel (et al) blow up with Meltdown and Spectre. It’s what they’d still be pushing if it weren’t for physics.

Performance and features. In single threaded performance, Apple's A chips are way ahead. That's because Apple highly customizes and optimizes their core designs, while Qualcom and Samsung use largely vanila ARM reference designs with maybe a few tweaks. Apple chips also have considerably more cache. That's difficult for the competition to counter because they are very price sensitive and cache size eats up expensive die space.

The counter argument is that the other systems make up for this by having more cores, which is really a cop out. Single threaded execution has a far more direct effect on user experience, gaming performance, etc. They can't compete in core engineering so make it up by slapping on more fairly generic cores. It does appear that Samsung is responding to this and investing in more advanced core designs.

The final area is custom features like the neural engine behind real-time face recognition, real time 3D face lighting effects and such. The competition don't really have these at all, so we don't know how far behind they are.

There's a pretty good article on this linked below.

https://seekingalpha.com/article/4138071-apple-cpu-advantage...

The Samsung M1, M2 and soon the M3 are all fully custom.

Yep, it does look like Samsung may be in a position to catch up soon. That could make a real problem for Qualcom the other Android manufacturers. Samsung are the biggest manufacturer of high end Android phones. If Qualcom can't sell high end CPUs to them any more, it might make manufacturing any truly high end CPUs uneconomical, or at least drive up costs even further.

Watch out for what comes out of ARM Austin. Their last core was the A72 from a couple of years ago. Rumour has it they've been working on a big core to better compete with Apple and Samsung.

Looking back, it makes sense that Qualcomm discarded their underperforming custom design efforts (at least for the mobile market). If ARM can deliver a competitive design, why not fully commit to their roadmap and save significant amounts on R&D costs?

Because if ARM cannot deliver they are fucked.

A lesson Apple learned at the hands of Motorola.

I sometimes wonder how much of It Just Works was influenced by having slower machines. When doing everything takes longer, if you do it right the first time then it’s still faster than doing it twice.

ARM is not just one company - it's a semi-standard.

Would you mind backing up your claim? To my knowledge, ARM is one company that licenses their designs to other companies, who are free to parameterize and modify as they please.

I don't know what the above comment was referring to but they are more than what you describe. I'd submit AMBA as an example of this; I've seen this tech utilised by IP for ASICs that don't actually contain any ARM processor cores themselves.

https://en.m.wikipedia.org/wiki/Advanced_Microcontroller_Bus...

I think the idea is that it probably matters very little to Apple or Samsung or any other company that has an architectural license how well the ARM originated core performs because they are building their own core and ARM is basically a standard committee to them. Obviously companies that are using the actual ARM cores care about ARM's ability to perform.

Why don't companies just come up with their own designs?

First off designing a whole core architecture from scratch is a huge and very expensive undertaking. Youd need to be able to get a really big competitive advantage to make it worthwhile.

Secondly theres a large ecosystem of add on components designed work with ARM and be dropped into ARM SOCS, such as GPUs, Wifi modules, Gyroscopes, wireless modems, GPS, etc.

Thirdly there are a lot of SOC engineers very familiar with ARM. You can hire them streight from competitors, or college, including PHDs that have done research on it. Youd need to train up any new hirs from scratch on your architecture.

Finally theres a huge software development tool chain built around the established processor architectures. To support a new architecture youd also need to build a set of compiler back ends, bearing in mind the existing ones benefit from many years of tweaking and optimization.

Yes, SW is the key; especially Linux. ARM is way ahead of other embedded CPU architectures in terms of Linux support.

There are two types of ARM licenses. One is an Architectectural license. This allows licensees to design and implement the ARM ISA. Apple and QCOM among others have Architectural licenses. They can add customizations such as # pipelines stages. This license is very expensive.

The other license is getting ARM's implementation of a CPU (ie A57, A72 etc). Customers can parametize # cores, cache size, memory width etc, but the basic architecture is fixed.

Because it gives you no competitive advantage.

Apple made the decision to pursue performance at the cost of device life.

>Performance and features.

I think both these metrics are now adequate, battery life is the single metric I am interested in now.

Battery life is, realistically, part of performance- it’s just “per watt”. Apple is also a leader in that area as well.

>That's because Apple highly customizes and optimizes their core designs, while Qualcom and Samsung use largely vanila ARM reference designs with maybe a few tweaks

Here's the Anandtech article on the "few" tweaks Samsung has made to its 6 wide decode custom M3 cores in the upcoming Exynos 9810.

https://www.anandtech.com/show/12361/samsung-exynos-m3-archi...

Apple's CPUs win benchmarks, but a lot of this is more down to economics. They're making CPUs for a single client, themselves, mostly for the high end, and they don't need to make a profit on them.

Qualcomm probably could make better CPUs...but it needs to sell them for a profit, and they might not fit into all phone's physical size or power budget.

So, Qualcomm ends up making somewhat lowest-common-denominator chips. Economies of scale make it difficult to make a run of super good chips when Samsung uses its own in many markets, Apple uses its own--the world of high-end smartphones outside of Samsung and Apple is just too small.

Isn't Apple cheating?

They're taking their CPUs and clocking them at speeds the device can't support, except when brand new, then letting the device slow down over the corse of its usable lifespan.

It would be interested to compare an Apple and Qualcomm CPU after a year and a half, to see how the benchmarks have changed.

That's incorrect. The CPUs don't degrade, the batteries do.

The CPU speed decreases. So doesn't matter if it technically is still able to reach higher clock speeds. In reality it runs slower.

Right, as an alternative to random crashes. Bear in mind this is not an Apple problem, its a battery problem. All the other manufacturers have this problem, they just dont attempt to detect battery power fluctuations and mitigate it with throttling. They just eat the system crashes. Its got nothing to do with the CPU design, otherwise replacing the battery wouldnt fix the problem.

> All the other manufacturers have this problem, they just don't attempt to detect battery power fluctuations and mitigate it with throttling. They just eat the system crashes.

Really, because I have not heard of any other manufacturer with this problem?

Then you have not looked into it. Android forums have been filled with complaints over “random crashes” when doing something CPU heavy for years.

Care to cite any sources as I'm only coming up evidence that they don't (the last one is a forum like you said and the reasoning explained there is due to bloat)

https://mashable.com/2018/01/04/apple-slowing-down-iphones-w...

https://www.theverge.com/circuitbreaker/2017/12/29/16829512/...

https://www.engadget.com/2017/12/29/samsung-lg-claim-older-p...

https://forums.androidcentral.com/samsung-galaxy-note-8/8589...

Source: basic understanding of how batteries & CPUs work while following tech forums for decades.

This isn’t considered controversial

You're completely ignoring the question. Do other smartphones start out fast due to new batteries and slow down due to the battery degradation like Apple or do they account for the battery degradation and keep performance stable?

Other smartphones either do the same thing (slow down) or just plain crash (most of them do this). It’s a common complaint.

It takes quite awhile for the battery to degrade to the point it can’t run the CPU at max capacity anymore. And it still could, but it would be potentially unstable.

"Right now Apple is several years ahead of other high end phone manufacturers in CPUs. If they cripple Qualcomm's high end modem business by taking most of it away from them, they could extend the same lead to modems."

Err, certainly this would just result in vertical integration (IE one of these companies buying Qualcomm), or any other situations, rather than "everyone sucks except for Apple" as you posit

Err, certainly this would just result in vertical integration (IE one of these companies buying Qualcomm), or any other situations, rather than "everyone sucks except for Apple" as you posit

Or, it could result in vertical integration and "everyone sucks except for Apple."

>Right now Apple is several years ahead of other high end phone manufacturers in CPUs

Does Apple even integrate a modem into their SoC's? I find it hard to believe that they can be years ahead of other high end phone SoC's manufactures when they haven't even accomplished that. Additionally, Samsung claims a 2x increase in their new Exynos 9810 SoC [1] which means that it should have single thread performance that is in the ballpark of Apple SoC's.

With a clock speed of up to 2.9GHz, a 3rd generation custom CPU offers higher computing power so that its single-core and multi-core performances are improved around two-fold and 40 percent respectively when compared to its predecessor.

[1]http://www.samsung.com/semiconductor/minisite/exynos/product...

They probably have milk in the fridge bought from a store even though they have had time to breed their own cows. They probably don't want the hassle of it when the whole herd could get foot and mouth disease. They let the milk supply chain deal with such things.

>The result: a chip powerhouse that could one day threaten the dominance of Qualcomm Inc. and even, eventually, Intel.

If by dominance being the largest Semiconductor company then yes, possibly. But not because Apple continue to grow infinitely, but both competitor are shrinking.

Apple is already close to Intel in terms of processor shipped, and is certainly larger then Intel in terms of CPU. And when Apple decide to have its own modem, Intel will ship another ~150M less unit.

Apple has roughly 20% of world wide market share. Qualcomm, with the shrinking % of 300M unit modem to Apple, may one day lost all of that when Apple decide to make it themselves. For the other 80%, there is roughly 20% at the bottom that Qualcomm will not compete in. With less then 60% of market, 40% of those are from Samsung, Huawei, BBK, and Xiaomi. Both Samsung and Huawei are planning to use their own SoC in the high end flagship phone, BBK are thinking of making their own as well. And you have Mediatek making gains at the bottom to mid end of the market. All of a sudden Qualcomm has less then 50% of Smartphone market to work with.

If Apple do actually have an team working on Modem, I wonder when will Apple becomes the largest purchaser in Semiconductor.

My understanding was that Samsung (Exynos) and Huawei (Kirin) do release global models of their phones based on their own SoC for GSM markets.

But they can't overlook the Nth American market where Snapdragon dominates the US with CDMA2000 support.

That is assuming CDMA 2000 is still relevant. You can wonder if anyone using a Flagship Smartphone cares about dropping CDMA support. The world has moved to LTE, and only much more so in the future.

CDMA matters less and less in the US, but for some reason Qualcomm is seemingly the only company making a modem with comprehensive US LTE band support.

And Intel isn’t? Intel’s modems have substantially all LTE bands.

Say what? Intel modems covers all US LTE bands that Qualcomm covers. Qualcomm has an edge on max bandwidth and supporting CDMA, but not LTE bands (for the US anyway).

Save for posing a major threat to suppliers who rely heavily on them, it would be a surprise if Apple sells its chips to anyone. Their chips are certainly great, if the responsiveness of their iOS devices is anything to go by. The OS and restrictions on background processes doesn’t harm them either. I wish they could somehow share their all advances with everyone.

This brings me to my main problem with Apple; as a consumer of technology in a the developing world. For Apple, I don’t exist, or at least the vast majority of my us don’t. The technology is always priced beyond our reach. Some individuals can afford them, but the vast majority have to use alternatives. Any success in slowing down Android would have meant no smart phones for us. Most people I know can’t afford an iPhone. This is what I consider the other side of the Apple story.

I don't think that's the 'other side' of Apple's story.

Respectfully, Apple is/was/tries to be a luxury and aspirational brand. They're expensive, everyone knows it. It's like saying that BMWs fault is that their cars aren't cheap.

I suppose the whole "luxury and inspirational" bit is my problem. Maybe an even bigger part of this is the fact that I don't look at technology the same way I look at cars. More and more, technology is becoming a necessity, more than a luxury. So I find myself routing against Apple, even if I love their phones and tablets. I hope this doesn't make me a hypocrite.

Totally unrelated; car manufacturers share technology quite a bit.

But cars are a necessity as well (for some parts of the world), and there's cars to fit all sorts of price points in the same way that you get get smartphones and tablets at any price.

They might have trouble selling their chips. They optimize far more towards performance over price than Qualcomm does. Apple can afford to do that, because of its high profit margins. Potential buyers probably couldn't afford that tradeoff.

The only market I can think of where higher prices wouldn't be as much of an issue is the server market. Imagine that. High density, Apple designed, ARM SOCs for servers. Or maybe they'll bring back Xserve.

But if you look at their chip design road map they started with an ARM Soc, now they have a motion co-processor, integrated graphics, and an ML chip. The road map of their SoC has been to specialize towards phone applications exclusively, the majority of that stuff just wouldn't make any sense in a server.

So what you're asking is not only for them to sell their IP to 3rd parties (a business model of partners they've repeatedly crushed) but also to build out an entirely separate chip design team focused on a new market.

On the other hand, servers are running more GPU / ML workloads these days and semiconductor designs are more automated. It’d be interesting to see if they could cost-effectively sell those designs with, perhaps, capacity tweaks similar to what they’ve done with the iPad X-series releases where the chip design is very similar but they’ve added extra GPU cores, cache, etc.

That’d also be consistent with their past history of trying to pick some edge for a particular market: optimized for media processing, ML, VR/AR, etc. rather than just competing with x86 for the generic workloads. I still don’t see it as a high probability but I’d expect a niche if it happens at all.

I just want Apple to team up with Nintendo and make a home console + Apple TV.

I’ve always felt Apple and Nintendo have a common thread that rarely gets appreciated.

That would be awesome, but it’s difficult to see how that would fit in with Nintendo’s vision, which seems to be merging console and mobile gaming. How do you do that without competing directly with iPhone/iPad?

Most of Nintendo's hardware obsession ends up being accessories more than core system tech. I would like to believe that Nintendo bringing ideas for hardware with Apple executing would work really well. Plus, Apple could use a bit more, "Friendly for children aesthetic" and not so much "Future Techno Glass World".

Would IMO totally make sense in markets where they cannot loose anything, like servers, workstations and PCs. Question is whether Apple has what it takes to disrupt completely new markets (with their current management).

I would suggest Apple has absolutely no interest in those markets. Even when Steve Jobs had been fired, Apple was focussed on products for the end-user, not components for others to assemble from.

Apple won't sell their components unless its MFi DRM chips. They aren't in the same market segment as Intel, because they don't sell CPUs for consumer or server markets.

Reply on news.ycombinator.com