M3 Chip Generation - Discussion Megathread

[leman]

From a marketing standpoint, it looks bad. Apple focused more on improving performance than improving efficiency at the keynote. If Apple had designed the A17 to improve efficiency, Apple should have focused on that during the keynote.

It's way to early to make these statements. We already have a precedent in A16, which sounded really lacklustre at the announcement, but ended up being significantly faster and more power-efficient that A15.

It seems that TSMC N3B increases performance by 10~15% with the same power compared to TSMC N5.

This hardly matters as we are talking about a new microarchitecture. We don't know how much of these 10% are IPC improvement and how much is the clock, nor do we know how the rest of the CPU comes together. Maybe it's clocked 10% lower than A16. We need more data.

 

[senttoschool]

It's way too really for this kind of pessimism. The crucial detail will be the clocks. If A17 runs at 3.6Ghz to achieve this 10% uptick, then it's indeed not a good sign. If instead it runs at 3.4Ghz, that would be an impressive IPC improvement. Shouldn't be too long until we have these details. My personal guess is that the architecture is capable of much higher clocks, they just wanted to limit the peak power consumption on the phone because it's already years ahead of closest competition.

Don't get infected by the likes of sunnyboy the fifth, that dude would still complain about "disappointing Apple" even if they released an Apple Watch that can take on an Nvidia GPU...

I'm not being pessimistic. Probably just realistic at this point.

I agree with @Xiao_Xi above. If Apple focused on efficiency, they would have boasted about it and maybe we'd see better battery life as well. But they didn't and the battery life remains the same.

Also, Apple would have configured the chip to yield the best efficiency/performance ratio. If 10% is all they got by moving from 5nm to 3nm, then it isn't that great.

I also wonder if this is just a die shrink of A16 CPU from 5nm to 3nm. That would explain the 10% because N3B is roughly 10% faster at the same power as 5nm.

 

[bobmans]

I'm not being pessimistic. Probably just realistic at this point.

I agree with @Xiao_Xi above. If Apple focused on efficiency, they would have boasted about it and maybe we'd see better battery life as well. But they didn't and the battery life remains the same.

Also, Apple would have configured the chip to yield the best efficiency/performance ratio. If 10% is all they got by moving from 5nm to 3nm, then it isn't that great.

I also wonder if this is just a die shrink of A16 CPU from 5nm to 3nm. That would explain the 10% because N3B is roughly 10% faster at the same power as 5nm.

I’m starting to think that this was nothing more than a die shrink with some minor or no changes to the CPU.

We’ll have to wait until the first benchmarks that list the clock speed though. If they increased the clock speed for these gains then I’ll be disappointed, if they kept the clock speed the same (or even lowered it), then I’ll be pleasantly surprised.

Battery life staying the same makes me think they increased clock speed rather than architectural improvements.

 

[senttoschool]

I’m starting to think that this was nothing more than a die shrink with some minor or no changes to the CPU and GPU cores (expect another core) + better NPU + USB controller + AV1 decoder.

We’ll have to wait until the first benchmarks that list the clock speed though. If they increased the clock speed for these gains then I’ll be disappointed, if they kept the clock speed the same (or even lowered it), then I’ll be pleasantly surprised.

Battery life staying the same makes me think they increased clock speed rather than architectural improvements.

The N5 to N3B speed improvement at the same power is 10%. Hence, it wouldn't shock me if it's just a die shrink of the A16.

Clearly, they improved the NPU and the GPU. Those seem to be new. But the CPU could just be a die shrink.

Perhaps N3B was an emergency node TSMC had to rush out in order to meet Apple's iPhone 15 timeline. If so, then maybe Apple decided to simply do a die shrink instead because they did not have enough time to do a complete redesign. Instead, the redesign would come using N3E.

 

[Xiao_Xi]

We’ll have to wait until the first benchmarks that list the clock speed though. If they increased the clock speed for these gains then I’ll be disappointed, if they kept the clock speed the same (or even lowered it), then I’ll be pleasantly surprised.

Battery life staying the same makes me think they increased clock speed rather than architectural improvements.

Apple could have avoided speculation about whether they face some problems by designing the CPU by explaining their goals and showing metrics that demonstrate they meet their goals. They have done this all along.

 

[gusping]

From a marketing standpoint, it looks bad. Apple focused more on improving performance than improving efficiency at the keynote. If Apple had designed the A17 to improve efficiency, Apple should have focused on that during the keynote.


It seems that TSMC N3B increases performance by 10~15% with the same power compared to TSMC N5.

	N3E vs N5	N3 vs N5
Speed Improvement @ Same Power	+18%	+10% ~ 15%
Power Reduction @ Same Speed	-34%	-25% ~ -30%
Logic Density	1.7x	1.6x
HVM Start	Q2/Q3 2023	H2 2022

TSMC Outlines 3nm Roadmap: Five FinFlex Nodes

TSMC's N3 node looks to have a big and bright future with FinFlex.

www.tomshardware.com

I agree with you. I'm shocked at how bad the A17 is on paper. 10% speed improvement on two cores. Nothing on the efficiency cores. No improved in battery life (granted many factors contributed to this).

I know it isn't comparable to a phone chip, but watch AMD release zen 5 with a 25% speed increase at the same power budget if the rumours are to be believed. What is going on at Apple? Lost all their best chip architects?

 

[leman]

I also wonder if this is just a die shrink of A16 CPU from 5nm to 3nm. That would explain the 10% because N3B is roughly 10% faster at the same power as 5nm.

They wouldn’t have mentioned a wider architecture if this were a die shrink.

 

[quarkysg]

Hence, it wouldn't shock me if it's just a die shrink of the A16.

Didn't Apple say that the A17 Pro was designed on a new micro-architecture with it being wider or something like that?

 

[senttoschool]

They wouldn’t have mentioned a wider architecture if this were a die shrink.

I missed that part. So maybe it's new but N3B is worse than expected or the new architecture is worse than expected.

 

[Xiao_Xi]

Didn't Apple say that the A17 Pro was designed on a new micro-architecture with it being wider or something like that?

 

[Natrium]

Apple didn’t even claim the a17 pro cpu (not gpu) is 10% faster. They only stated the performance cores are up to 10% faster. There’s no performance gain for the efficiency cores, which according to Apple are used “most of the time”. In other words, most of the time there’s no performance difference compared to the a16, and only some of the time there’s a meagre up to 10% performance gain.

 

[gusping]

Apple didn’t even claim the a17 pro cpu (not gpu) is 10% faster. They only stated the performance cores are up to 10% faster. There’s no performance gain for the efficiency cores, which according to Apple are used “most of the time”. In other words, most of the time there’s no performance difference compared to the a16, and only some of the time there’s a meagre up to 10% performance gain.

Surely something went wrong during the SoC design or manufacturing phase. This is hilariously bad for an N3 process.

 

[senttoschool]

Apple didn’t even claim the a17 pro cpu (not gpu) is 10% faster. They only stated the performance cores are up to 10% faster. There’s no performance gain for the efficiency cores, which according to Apple are used “most of the time”. In other words, most of the time there’s no performance difference compared to the a16, and only some of the time there’s a meagre up to 10% performance gain.

It doesn't matter much for M3 because we care mostly about the performance cores.

 

[souko]

It can be that Apple no longer thinks that saying how much better they are is sexy. (In the world of AI they said, that their Neural engine is twice the performance.) This is consumer conference. Let´s wait for in depth reviews. Even at WWDC Apple did not speak about foundation framework and so on. And iPhone event is focused for consumers much more than WWDC

 

[Adult80HD]

I find it amusing how much hand-wringing there is over performance with no data to even judge. Let's face it, the iPhone is already by leaps and bounds the fastest mobile device you can buy. It's not like we need more speed here for almost any tasks. The one area it could have been improved is the GPU/graphics for gaming, and they've done that. We know in the past they left it out because it used too much power on the previous generation chip. IMO the logical conclusions here is that Apple has approached this as a power budget issue: Without increasing battery size/weight, where can we use our power/wattage budget to give users the best performance boost they care about? The new GPU is a highly logical approach.

I wouldn't extend many conclusions from this to the desktop/laptop versions of the M3; Apple can a lot of different things there without the power and cooling constraints.

 

[huge_apple_fangirl]

Surely something went wrong during the SoC design or manufacturing phase. This is hilariously bad for an N3 process.

Well Apple lost a bunch of top chip designers a few years ago. That could just start to be showing up now. Also WFH probably lowered productivity of the chip team that also wouldn't show up right away. And process improvements on the manufacturing side have massively slowed down and 3nm isn't a huge improvement.

In retrospect, Apple really chose the absolute best time to ditch Intel. The gap between them and Intel in late 2020 in both design and process (Firestorm vs Tiger Lake, TSMC 5nm vs Intel 10nm (now Intel 7)) is probably the largest it's ever going to be. IPC improvements on Apple's side have massively slowed down since then, as has TSMC's process advancements.

 

[leman]

I find it amusing how much hand-wringing there is over performance with no data to even judge.

Precisely. The worries appear to be based around the arguments like "but they used to boast with 20% better performance in the past!", and while I sympathise with those posters, these are not good arguments. Times change, as do design goals.

For example, Apple has very little incentive to make the new phone CPU 20% faster (they are already years ahead of the competition), but they have incentive to make Mac chips faster. Maybe this new CPU is designed with higher power draw and higher clocks in mind (unlike A14/A15 that pretty much peaked at 5W per core max). Then Apple would need to clock it more conservatively in the iPhone, so that they can meet their power targets.

Of course, what I wrote above is just a speculation, and might as well be baseless, but I hope it illustrates that there can be plenty explanations for only modest performance improvements in A17, not all of them equally gloomy. We will know more when we have more data bout the chip (performance, clocks, features, etc.). Then we will have at least something to speculate about Apple's inevitable impending doom.

 

[dgdosen]

I don't think people here are hand-wringing versus just trying to read into Apple's opaque disclosures about their SoCs. I like reading these opinions - educated and not - and imagine we'll see perf numbers closely before or on launch date.

So my latest open questions are:
- Will M3 CPU cores be based on the A16? A17(Pro)? something newer?
- will M3 GPU cores be based on the A16? A17(Pro)?
- With forecasts of no M3 (laptops) this year - will M3s be built on N3B or N3E?
- Did N3E even start volume production yet?
- At what speeds will A17 and M3 run?
- Is Apple really "not leaving anything on the table" wrt perf?

 

[leman]

@dgdosen If M3 is based on A16, it would be shocking and sad.

 

[gusping]

Precisely. The worries appear to be based around the arguments like "but they used to boast with 20% better performance in the past!", and while I sympathise with those posters, these are not good arguments. Times change, as do design goals.

For example, Apple has very little incentive to make the new phone CPU 20% faster (they are already years ahead of the competition), but they have incentive to make Mac chips faster. Maybe this new CPU is designed with higher power draw and higher clocks in mind (unlike A14/A15 that pretty much peaked at 5W per core max). Then Apple would need to clock it more conservatively in the iPhone, so that they can meet their power targets.

Of course, what I wrote above is just a speculation, and might as well be baseless, but I hope it illustrates that there can be plenty explanations for only modest performance improvements in A17, not all of them equally gloomy. We will know more when we have more data bout the chip (performance, clocks, features, etc.). Then we will have at least something to speculate about Apple's inevitable impending doom.

I agree it is nothing to worry about. But... if there were large performance gains, Apple would shout about them. First party figures/benchmarks are almost always the highest figure given (when compared to independent reviews). If they were going all in on efficiency, why has battery life stayed the same (unless the battery size itself has shrunk massively). It all seems odd to me. I just hope the M3 is at least 20% faster.

 

[leman]

I agree it is nothing to worry about. But... if there were large performance gains, Apple would shout about them.

I fully agree. It’s just them not shouting about it doesn’t mean that there are no gains. We already saw this with A16. Maybe they are changing their marketing strategy. Who knows.

 

[name99]

I’m starting to think that this was nothing more than a die shrink with some minor or no changes to the CPU.

We’ll have to wait until the first benchmarks that list the clock speed though. If they increased the clock speed for these gains then I’ll be disappointed, if they kept the clock speed the same (or even lowered it), then I’ll be pleasantly surprised.

Battery life staying the same makes me think they increased clock speed rather than architectural improvements.

They specifically called out on a slide that there are changes to branch prediction (and we have a pretty good idea what those are) and increases to decode and execution width.
Point is not that these are the only changes, but that these are changes that you don't get from "just a die shrink".


Battery life is now "problematically" correlated with better CPU/GPU efficiency. Look at what Apple tells you on the Tech Specs web page:
Video playback (in both versions) depends on screen, media block, and network – NOT on CPU or GPU. Same for Audio playback.
They matter (insofar as they are useful info to people, eg going on a flight) but they are NOT for example
- idle time (how long does your phone last with minimal use?)
- photo time (on vacation, taking photos every few minutes)
- game time (heavy CPU and GPU usage)
- web browsing time (mostly heavy CPU usage)
- substantial dictation time (lots of NPU [and maybe CPU and GPU?] usage)

etc etc
I suspect Apple has stopped talking about battery lifetime not because it's not important but because there's no way to do so that's a win for them. They can talk up how one use case is much improved (say dictation) and there's a whole crowd on the internet that will immediately say "Apple lies about battery life" because some very different use case (playing video?) is unchanged...

 

[name99]

-

 

[name99]

I fully agree. It’s just them not shouting about it doesn’t mean that there are no gains. We already saw this with A16. Maybe they are changing their marketing strategy. Who knows.

Don't forget that a certain idiot crowd of politicians and lawyers have now decided that Apple's business is their business... (And the USB-C change has clearly made this very salient to Apple).
Which means that ANY specific number Apple announces about performance or battery life has the problem that some idiot is going to insist that it was deliberate deception because blah blah blah and therefore Apple owes them a million dollars or needs to be broken up.

There's really no way Apple can win, going forward, by saying anything but the most banal statements, and letting reviewers fill in the numbers.

 

[Xiao_Xi]

It’s just them not shouting about it doesn’t mean that there are no gains.

I suspect Apple has stopped talking about battery lifetime not because it's not important but because there's no way to do so that's a win for them.

Not true. Apple shows the improved CPU and battery life on the website.

certain idiot crowd of politicians and lawyers have now decided that Apple's business is their business

You should be more respectful of elected politicians. Just because they don't represent you doesn't mean they don't represent others.