Discussion Apple Silicon SoC thread

[Eug]

M1
5 nm
Unified memory architecture - LP-DDR4
16 billion transistors

8-core CPU

4 high-performance cores
192 KB instruction cache
128 KB data cache
Shared 12 MB L2 cache

4 high-efficiency cores
128 KB instruction cache
64 KB data cache
Shared 4 MB L2 cache
(Apple claims the 4 high-effiency cores alone perform like a dual-core Intel MacBook Air)

8-core iGPU (but there is a 7-core variant, likely with one inactive core)
128 execution units
Up to 24576 concurrent threads
2.6 Teraflops
82 Gigatexels/s
41 gigapixels/s

16-core neural engine
Secure Enclave
USB 4

Products:
$999 ($899 edu) 13" MacBook Air (fanless) - 18 hour video playback battery life
$699 Mac mini (with fan)
$1299 ($1199 edu) 13" MacBook Pro (with fan) - 20 hour video playback battery life

Memory options 8 GB and 16 GB. No 32 GB option (unless you go Intel).

It should be noted that the M1 chip in these three Macs is the same (aside from GPU core number). Basically, Apple is taking the same approach which these chips as they do the iPhones and iPads. Just one SKU (excluding the X variants), which is the same across all iDevices (aside from maybe slight clock speed differences occasionally).

EDIT:



M1 Pro 8-core CPU (6+2), 14-core GPU
M1 Pro 10-core CPU (8+2), 14-core GPU
M1 Pro 10-core CPU (8+2), 16-core GPU
M1 Max 10-core CPU (8+2), 24-core GPU
M1 Max 10-core CPU (8+2), 32-core GPU

M1 Pro and M1 Max discussion here:

Page 78 - Discussion - Apple Silicon SoC thread

Page 78 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M1 Ultra discussion here:

Page 109 - Discussion - Apple Silicon SoC thread

Page 109 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M2 discussion here:

Page 127 - Discussion - Apple Silicon SoC thread

Page 127 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

Second Generation 5 nm
Unified memory architecture - LPDDR5, up to 24 GB and 100 GB/s
20 billion transistors

8-core CPU

4 high-performance cores
192 KB instruction cache
128 KB data cache
Shared 16 MB L2 cache

4 high-efficiency cores
128 KB instruction cache
64 KB data cache
Shared 4 MB L2 cache

10-core iGPU (but there is an 8-core variant)
3.6 Teraflops

16-core neural engine
Secure Enclave
USB 4

Hardware acceleration for 8K h.264, h.264, ProRes

M3 Family discussion here:

Page 215 - Discussion - Apple Silicon SoC thread

Page 215 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M4 Family discussion here:

Page 263 - Discussion - Apple Silicon SoC thread

Page 263 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

 

[repoman27]

What do you make of the claim that M4 uses 50% of the power of M2 at the same speed as M2?

Seems legit. M2 was Avalanche + Blizzard on TSMC N5P, while M3 was Everest + Sawtooth on N3. So you have µarch plus process improvements with that generation. Apple claimed the Everest performance cores in the M3 were up to 15% faster than Avalanche in the M2, and the Sawtooth efficiency cores were up to 30% faster than Blizzard. They also made the claim then that:

Together, these cores create a CPU that delivers the same multithreaded performance as M1 using as little as half the power, and up to 35 percent more performance at peak power.

Probably a similar deal with the comparison between M4 and M2, I'm sure you can imagine what the marketing slide comparing the curves would look like. But the M4 has 50% more of the cores that were 30% faster and gets another process improvement good for +5% speed at the same power or -7% power at the same speed.

Based on the 'best' scores in GB6 ST Apple gained about 15% from M2 to M3. Apple E cores tend to be worth about 1/3 of a P core, so an 8 core M2 had 5.3 cores vs 10 core M4 having 6 cores, accounting for 12%. From N3E TSMC says you should get 6-7%. If you multiply all that together (can't add it, since they are cumulative) you get 38%. They need to gain another 9% from other sources to achieve a 50% overall gain.

There is more memory bandwidth, that could help some MT tests that are memory bound but another mid single digit ST gain seems likely. That doesn't have to mean a new core, it could be essentially the same core with some minor reworking to improve cache size/latency, or some tweaking with FinFlex to use higher power/faster transistors in a few chosen places that are restricting the ability to up the clock a bit.

Another possibility is that the E cores saw some significant performance gains while the P cores remained pretty much flat, but it is harder to track down that information (it would be nice if Geekbench had a setting where it ran benchmarks pinned to a CPU's smaller cores...have to remember to suggest this next time John pops up on RWT)

The short turnaround between M3 and M4 makes it less likely we'd see any major changes. That becomes even more true if this "early" release of M4 was a one off and M5 is back to the fall. I think Apple has good reasons to want Apple Silicon to hit earlier in the year in general, but we'll have to see what happens next year to know.

Apple's not talking about single-threaded performance, they're talking about peak multi-threaded. M3 already beat M2 by more than 25% in that scenario because the e-cores saw more of an uplift than the p-cores, and the M4 has two more of them and all of the cores can clock higher.

Yeah that makes me think it’s basically artificial and more about segmentation here. The odds the one E core is that big of a deal yield wise is really low I think.

Apple doesn't bin for artificial market segmentation. Compared to other semiconductor companies, they barely bin at all. Apple has always quietly binned their SoCs based on power / leakage. Disabling cores is 100% about yields. In this case the lead products are iPads: passively cooled, ultra-thin tablets that have serious constraints when it comes to skin temperatures. Disabling a single performance core makes practically zero difference to the vast majority of end users, but probably allows them to salvage a massive number of chips that would otherwise not have met power / thermal targets when all of the cores were loaded. If they had started the M4 transition with the MacBook Pro and Mac mini, they might have been able to bin low leakage chips until they had enough to use in the iPad Pros. On the other hand, having led with the iPads, the M4s destined for future Mac minis and MacBook Airs may have all their CPU cores enabled but instead have one or two GPU cores disabled, allowing Apple to salvage yet more dies for use in products where GPU performance is less critical.

The person I quoted claims that M4 uses the same core design as M3. Therefore, there is no new core claimed. Furthermore, the node change likely doesn’t help much in performance. Some day N3B is faster while others claim N3E is faster.

So I was asking the person who made those claims for an opinion on how M4 is twice as efficient as M2 while Apple did not make the same claim for M3.

I already responded to your initial question, but I just wanted to address the bit about process improvements. The only claims that matter are the ones TSMC is making, and as I have pointed out several times in this thread now, TSMC says N3 to N3E is +5% speed at the same power or -7% power at the same speed. Random people spew nonsense on the internet all the time, but if a credible source with first-hand knowledge of TSMC's N3 processes has publicly disputed TSMC's performance claims, feel free to provide links.

 

[repoman27]

Interesting. The Apple logo on the back of the new iPad Pro is actually a copper heatsink.

Yeah, I like how Apple had to get out ahead of the people doing teardowns who will go, "Derp, Apple's new $3000 iPad Pro doesn't even have a heat sink!" by explaining their thermal solution during the product launch. iFixit seems to have no idea that the black film that Apple has been using for a while now is a graphite / carbon based heat spreader. I wonder how many FLIR images of the Apple logo on the new iPad Pros will get posted as a result of this.

 

[Joe NYC]

No way I'm getting Nanotexture

Not only is it only available on the 1 & 2 TB models, I really don't like the look. Also, they left the bezels glossy so it's a sharp transition from the bezel to the screen. I guess they left the bezels glossy for the FaceTime camera.

Anyhow, still nothing on Geekbench, but like I said, given that machines aren't out until next week, it may be a couple of days before we see some Geekbench leaks.

I have seen this image all over Twitter today, wondering if it is the TV Dinner Tray from Fallout 4, and how much Aluminum it might have...

 

[trivik12]

https://twitter.com/x/status/1788010620509687999

Is this 1st benchmark with new iPad Pro. Not sure if there are any comps.

This is M3 Pro and that is slightly faster. So impressive number for M4.

 

[Doug S]

Apple is saying "up to" so they could have picked something that benchmarks a bit better for whatever reason.

I remember everyone thought the same thing when they announced a 70% ST gain for A9, then it turned out to actually be 70% faster in real life.

A 50% MT gain is trivial, I did the math for it a few posts ago. They only need a single digit IPC gain on the P core, or a low double digit gain on the E core, to make that happen. You're acting as though they're making some crazy claim that could only be true if they're being shady with their benchmark choices.

Yet over in the AMD Zen 5 thread plenty of people are believing a 40% IPC gain is coming, and a few seem to think AMD is lowballing that.

Anandtech forums can be weird sometimes!

 

[Doug S]

As for why Apple didn't make the same claim for N3: the gains were likely even smaller for M3 vs M2, so Apple likely just didn't want to talk about it.

The reason why they did the compare to M2 instead of M3 is blindingly obvious to anyone who is paying attention. They were announcing iPad Pros, and the previous iPad Pro contained an M2! Why would they compare with the M3, when there never was an M3 iPad Pro?

When they do a new Mac Pro or Mac Studio with M4 Ultra they will compare with M2 Ultra for the same reason. When they do a new Macbook, if they compare with M2 instead of M3 then your argument might hold water, since there is an M3 Macbook.

 

[uzzi38]

The reason why they did the compare to M2 instead of M3 is blindingly obvious to anyone who is paying attention. They were announcing iPad Pros, and the previous iPad Pro contained an M2! Why would they compare with the M3, when there never was an M3 iPad Pro?

When they do a new Mac Pro or Mac Studio with M4 Ultra they will compare with M2 Ultra for the same reason. When they do a new Macbook, if they compare with M2 instead of M3 then your argument might hold water, since there is an M3 Macbook.

The question wasn't why did they compare M4 to M2, the question was why didn't Apple make the same claim for M3 vs M2. Look:

So I was asking the person who made those claims for an opinion on how M4 is twice as efficient as M2 while Apple did not make the same claim for M3.

 

[SpudLobby]

So: do we think it’s largely the same cores and the wider execution and decode they claimed is too just in reference to the M2 cores — just as the 50% was in reference to the M2?


I’m sure the cores have nonzero physical design tweaks vs the M3, FinFlex and new design rules probably encourage that. But more interested in if they have even small arch changes as indicated (over the M3).

 

[SteinFG]

Right now, apple is selling
M1 (5nm) : None
M2 (5nm) : Macbook Air, Mac Mini, Mac Studio, Mac Pro, ipad Air.
M3 (3nm) : Macbook Air, Macbook Pro, iMac.
M4 (3nm) : ipad Pro.

My guess is that M2 and M4 will stay for a bit longer, while M3 will get phased out.

 

[SteinFG]

So about that N3B node, it seems like apple really wants to jump off of it. Then I guess A17 Pro won't be coming to the next regular iphone 16. Here's my guess for how iPhone 16 will turn out

iphone 15	iphone 16	iphone 16 pro
A16 (N5, Fully enabled chip)	A17 (N3E, cut down A18 Pro)	A18 Pro (N3E, Fully enabled chip)
16 bil. transistors	~20 bil. transistors	~20 bil. transistors
6 GB	6 GB	8 GB
$700	$800	$1000

 

[Gideon]

I remember everyone thought the same thing when they announced a 70% ST gain for A9, then it turned out to actually be 70% faster in real life.

A 50% MT gain is trivial, I did the math for it a few posts ago. They only need a single digit IPC gain on the P core, or a low double digit gain on the E core, to make that happen. You're acting as though they're making some crazy claim that could only be true if they're being shady with their benchmark choices.

Yet over in the AMD Zen 5 thread plenty of people are believing a 40% IPC gain is coming, and a few seem to think AMD is lowballing that.

Anandtech forums can be weird sometimes!

I'd add the Intel thread, where there seems to be a consensus (based on a blurred picture) that the next big core adds 50-100% more resources to every backend feature, has gazillion-wide decode capability (with ucache), yet only delivers <=20% IPC uplift

 

[Eug]

Here is AnandTech’s article. It a nice summary of everything that is known so far about M4.

Apple Announces M4 SoC: Latest and Greatest Starts on 2024 iPad Pro

www.anandtech.com

 

[junjie1475]

M4 still uses the same CPU microarchitecture as M3 same as well for the GPU.

 

[Muadib]

It's a definite buy for me. I wanted OLED, landscape camera, 256 GB base storage, AV1 (just because), and a lighter Magic Keyboard. I got all of the above. The price went up a bit more than I was hoping, but that increase was way, way lower than the crazy $500 prediction from some of the doom-and-gloomers. I was hoping for a US$100-150 price increase, but it went up $200, from $799 base to $999 base at retail (but with double the base storage).

Canadian edu pricing:

M2 iPad Pro
128 GB CA$1099 (US$800)
256 GB CA$1229 (US$895)

M4 iPad Pro
256 GB CA$1249 (US$909)

Also, next month we should get a CA$100 (US$73) gift card back.

Sounds good to me! I think I'll wait a bit too, just in case of an early sale. I want the 13 inch model, but haven't decided on the ram yet. If this thing can do ray tracing well, then I'll have to try a few games on it. I didn't care for the gaming experience on my current 11 inch pro.

 

[repoman27]

Here is AnandTech’s article. It a nice summary of everything that is known so far about M4.

Apple Announces M4 SoC: Latest and Greatest Starts on 2024 iPad Pro

www.anandtech.com

View attachment 98582

There's a typo in the chart there, the memory should be LPDDR5X-7500 not 7700. 128 x 7500 / 8000 = 120 GB/s.

edit: I realize it's Ryan's chart, not yours.

 

[Eug]

There's a typo in the chart there, the memory should be LPDDR5X-7500 not 7700. 128 x 7500 / 8000 = 120 GB/s.

I've seen a couple of articles now say 7700. Not sure where this number is coming from. Maybe they copied AnandTech?

https://mp.weixin.qq.com/s/K7pTcNIgOJb8TVgnVowkOw

[News] Apple Unveiled M4 Chip for AI, Heralding a New Era of AI PC | TrendForce Insights

On May 7 (The US time), Apple launched its latest self-developed computer chip, M4, which is integrated into the new iPad Pro as its debut platform. M...

www.trendforce.com

 

[repoman27]

So: do we think it’s largely the same cores and the wider execution and decode they claimed is too just in reference to the M2 cores — just as the 50% was in reference to the M2?


I’m sure the cores have nonzero physical design tweaks vs the M3, FinFlex and new design rules probably encourage that. But more interested in if they have even small arch changes as indicated (over the M3).

Interestingly, this is the first M-series launch where Apple hasn't given us a die shot, so we can't even compare the cores visually.

So about that N3B node, it seems like apple really wants to jump off of it. Then I guess A17 Pro won't be coming to the next regular iphone 16. Here's my guess for how iPhone 16 will turn out

iphone 15	iphone 16	iphone 16 pro
A16 (N5, Fully enabled chip)	A17 (N3E, cut down A18 Pro)	A18 Pro (N3E, Fully enabled chip)
16 bil. transistors	~20 bil. transistors	~20 bil. transistors
6 GB	6 GB	8 GB
$700	$800	$1000

It's been the rumor for a while now that all of the iPhones 16 will be getting the A18.

Apple is doing a tick-tick-tock to navigate the N3 transition.

iPhones 14 Pro (2022): A16 Bionic = Everest + Sawtooth on TSMC N4
iPhones 15 Pro (2023): A17 Pro = Everest + Sawtooth + new GPU + new NPU on TSMC N3 = tick
Macs (2023): M3 = Everest + Sawtooth + new GPU on TSMC N3 = tick
iPads Pro (2024): M4 = Everest + Sawtooth + new GPU + new NPU on TSMC N3E = tick
iPhones 16 (2024): A18 = new p-core + new e-core on TSMC N3E/N3S = tock

 

[Nothingness]

M4 still uses the same CPU microarchitecture as M3 same as well for the GPU.

Do you have a source for that? I didn't find anything except (educated) guesses.

 

[repoman27]

https://twitter.com/x/status/1788010620509687999

Is this 1st benchmark with new iPad Pro. Not sure if there are any comps.

This is M3 Pro and that is slightly faster. So impressive number for M4.

Here's that M4 iPad result vs. the top scoring M3 iMac with 16GB of RAM (I figure the 24-inch iMac is essentially a big iPad with a stand anyway): https://browser.geekbench.com/ml/v0/inference/compare/364912?baseline=372975

Interestingly, the CPU speed for that M4 result is only 3.93 GHz vs 4.05 GHz for the M3. So maybe Apple backed off a little on the max clock speed for the iPads?

 

[Mopetar]

Do you have a source for that? I didn't find anything except (educated) guesses.

No one has any knowledge one way or the other yet, but it does seem the most likely case.

 

[Nothingness]

No one has any knowledge one way or the other yet, but it does seem the most likely case.

That's what I think too, but the OP seems so unequivocal I thought he found something I missed.

 

[repoman27]

I've seen a couple of articles now say 7700. Not sure where this number is coming from. Maybe they copied AnandTech?

https://mp.weixin.qq.com/s/K7pTcNIgOJb8TVgnVowkOw

[News] Apple Unveiled M4 Chip for AI, Heralding a New Era of AI PC | TrendForce Insights

On May 7 (The US time), Apple launched its latest self-developed computer chip, M4, which is integrated into the new iPad Pro as its debut platform. M...

www.trendforce.com

Looks like TrendForce / DRAMeXchange (part of TrendForce) published slightly ahead of AnandTech, so maybe that's where Ryan got the number from? But they also said the memory clock speed of M4 only reaches about or approximately 7700 MT/s. Which is odd because 7500 is a bog standard bin for LPDDR5X and it exactly matches Apple's memory bandwidth figure.

 

[repoman27]

That's why I think too, but the OP seems so unequivocal I thought he found something I missed.

If you're referring to me, I don't know for certain, and none of us probably will until the folks that are willing to dig into stuff like that have devices in hand. But the probability is high enough that I'm willing to take it for granted at this point.

edit: Never mind, I see you were referring to @junjie1475.

 

[Mopetar]

That's why I think too, but the OP seems so unequivocal I thought he found something I missed.

I (and I suspect this is true for most others as well) would be extremely surprised if it weren't essentially the same core. Apple no doubt made some fixes and tweaks as a part of porting the design to N3E, but I don't think there's a fundamental difference between them M3 and M4 cores.

Although Apple is usually pretty mum about details, typically they will at least mention that they have a new core when introducing a new chip and provide some cherry picked benchmark to claim that it's XX% better at "important task" than the previous chip.

 

[Nothingness]

I (and I suspect this is true for most others as well) would be extremely surprised if it weren't essentially the same core. Apple no doubt made some fixes and tweaks as a part of porting the design to N3E, but I don't think there's a fundamental difference between them M3 and M4 cores.

Although Apple is usually pretty mum about details, typically they will at least mention that they have a new core when introducing a new chip and provide some cherry picked benchmark to claim that it's XX% better at "important task" than the previous chip.

I made a typo in my previous answer 'why' instead of 'what'. So I do agree. And I think @junjie1475 either affirms things without any source, or has a source we don't know about.