Discussion Apple Silicon SoC thread

Page 149 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
23,583
996
126
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:

Screen-Shot-2021-10-18-at-1.20.47-PM.jpg

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:


M1 Ultra discussion here:


M2 discussion here:


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:

 
Last edited:

Eug

Lifer
Mar 11, 2000
23,583
996
126
A16
16 billion transistors
4 nm
6-core CPU (2 perf, 4 eff) = "Faster" but doesn't say how much, 20% lower power
16-core neural engine - 17 trillion operations per second
5-core GPU - 50% more memory bandwidth

A16.jpg

So it seems A16 isn't much faster than A15 (for CPU).
 
Last edited:

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
So 16 billion transistors compared to the A15's 15 billion, so they probably cost exactly the same. But the A16 uses LPDDR5, which is probably responsible for the cost difference and the Pro / non-Pro separation.
 

Eug

Lifer
Mar 11, 2000
23,583
996
126
So 16 billion transistors compared to the A15's 15 billion, so they probably cost exactly the same. But the A16 uses LPDDR5, which is probably responsible for the cost difference and the Pro / non-Pro separation.
I wonder how 4 nm factors into the cost here.

So have we reached the end of the big speed jumps?
 

jpiniero

Lifer
Oct 1, 2010
14,509
5,159
136
So have we reached the end of big speed jumps?

It kind of seems like it? Maybe N4 wafers are more expensive than perhaps thought, but I'm wondering if we're seeing the impact of the lockdowns/WFH on Apple's productivity and they decided to just do a mostly simple shrink of A15 to N4 and not be adventurous. Kind of makes you wonder if the cpu design team is a complete mess right now.
 

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
Nope Apple is focused on power efficiency. Plus no major CPU redesign.

Yeah I wonder whether A16 is another fairly incremental update like A15, which other than a few tweaks for M2 like more address bits and some cache changes did not appear to change the big core much at all between A14 and A15, though the little core got a major update. The extra billion transistors may be primarily due to bigger ISPs, improved NPU, and LPDDR5 (those controllers are bigger and I guess Apple chose not to make them 4X compatible thus why only the Pro got the A16)

I had been assuming for a long time Apple would launch M2 with A16's cores, and was surprised it didn't. Though in hindsight, if A16 cores are basically the same as A15 what would be the point?

One would assume A17 will be a pretty major update if it is following two years of very minor changes while they concentrated their top resources on getting Apple Silicon in Macs all the way through Mac Pro, and N3 will provide a much larger transistor budget. Maybe we'll get a preview of it if they release M3 Macs next spring/summer.
 
  • Like
Reactions: Tlh97

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
It kind of seems like it? Maybe N4 wafers are more expensive than perhaps thought, but I'm wondering if we're seeing the impact of the lockdowns/WFH on Apple's productivity and they decided to just do a mostly simple shrink of A15 to N4 and not be adventurous. Kind of makes you wonder if the cpu design team is a complete mess right now.

N4 was designed to be cost optimized, those wafers will cost less than N5 for the same design getting the ~6% shrink and at most be at cost parity with the same sized die.

I think the A14/A15/A16 rehashes have more to do with all the work needed for Apple Silicon - which at best is only recently complete vis a vis the Mac Pro. Though it is possible their CPU design team is a mess - if A17 isn't a pretty big jump that's when we might need to worry.
 
  • Like
Reactions: Tlh97

uzzi38

Platinum Member
Oct 16, 2019
2,565
5,570
146
I wonder how 4 nm factors into the cost here.

So have we reached the end of the big speed jumps?
N4 is an ever so slight shrink to N5 of 6% higher logic density with no performance gains. It's designed to be cheaper than N5, not more expensive.

Node alone has never been the sole contributor to performance/efficiency gains, and we've known that from N5 onwards node was always going to play a smaller and smaller role to performance gains.

But Apple's slowdown over the last couple of years hasn't really been in line with the rest of the industry... it's been a lot more drastic since A14.
 
  • Like
Reactions: Tlh97

LightningZ71

Golden Member
Mar 10, 2017
1,627
1,898
136
Apple had two major events that seem to have taken a toll on their HP core development. The first was the amount of design and development resources that were required to be pumped into the M series processors to get that ecosystem started and expanded. The second was the loss of several of their lead engineers when they broke away to form Nuvia. Given what we know about the lead times on chip development, it's not shocking that designs that are being released years later seem to be suffering from a SEEMING lack of innovation on the performance front as compared to the developments by competitors. However, we can not overlook the simple fact that Apple had a significant lead in performance over the rest of the field several years ago and likely took the gamble that this likely inevitable slowdown in their own performance improvements would likely still have them in the lead over everyone else (which, arguably, they still are) would be worth freeing themselves from reliance on Intel designs.

So far, it hasn't hurt them. I do think that they have to find a way to continue to carry performance forward a bit more each year as the competition, especially in the form of the X1/X2 ARM cores and Google's Tensor SoCs, is certainly picking up some steam. It also seems to me that we're reaching a point across the board in mobile SoCs that HP core performance is no longer the hold up it once was. Most HP cores since the 710/x1/A13 are plenty fast enough for day to day phone operations. It's the rest of the SoC and clever programming that are going to push the platform forward.
 

gdansk

Golden Member
Feb 8, 2011
1,973
2,352
136
I'd just like to note that it's disappointing when Apple does 10% improvement each year but somehow amazing when Intel or AMD do a 21% improvement every two years.

Not that those are the real numbers but because they have been shipping a new A chip every year the gains will appear less dramatic.
 

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
I'd just like to note that it's disappointing when Apple does 10% improvement each year but somehow amazing when Intel or AMD do a 21% improvement every two years.

Not that those are the real numbers but because they have been shipping a new A chip every year the gains will appear less dramatic.

Also because people grew spoiled by some really big yearly jumps - I remember when they claimed 70% gain from A8 to A9 and a lot of people were laughing. Until the benchmarks came in and it turned out they weren't fudging. That was when people started taking the idea that Apple will someday switch from x86 a lot more seriously.

Obviously big jumps are a lot easier when you are catching up than when you are caught up.
 

Eug

Lifer
Mar 11, 2000
23,583
996
126
I'd just like to note that it's disappointing when Apple does 10% improvement each year but somehow amazing when Intel or AMD do a 21% improvement every two years.

Not that those are the real numbers but because they have been shipping a new A chip every year the gains will appear less dramatic.
Well, I am now guessing CPU is less than 5% faster this time around.

Not that it really matters. My iPhone 12 Pro Max with A14 is lightning fast. I'm more interested in the other chip elements that make the stuff like improved cameras possible.

My wife has an iPhone XR and wants an iPhone upgrade. I was planning to get her the iPhone 14 Plus (since as mentioned, I don't care about CPU speed, and neither does she), but the fact that they didn't increase the price on the iPhone 14 Pro Max makes that idea more attractive than the rumours were suggesting. The idea of a physically much larger and 48 MP image sensor outputting 12 MP images (with similar image size) is a definite draw. Presumably the improved image processor of A16 was an absolute requirement to make that happen.
 
  • Like
Reactions: Tlh97

Eug

Lifer
Mar 11, 2000
23,583
996
126

Nothingness

Platinum Member
Jul 3, 2013
2,371
713
136

poke01

Senior member
Mar 8, 2022
651
612
106
I guess the arguments about M2 being based off the A15 vs A16 CPU core may be moot, since they might just be the exact same thing.
They are the same but A16 has less power usage due to 4nm. M2 is 3.49Ghz and has around 1930 single core and A16 is clocked at 3.46Ghz and gets 1880. I think the A16 would also get 1930 but its throttled in the iPhone.

Nuvia cores need to arrive fast Apple is slacking too much.
 
  • Like
Reactions: Tlh97 and Racan

Lodix

Senior member
Jun 24, 2016
340
116
116
I guess the arguments about M2 being based off the A15 vs A16 CPU core may be moot, since they might just be the exact same thing.
Taking in mind that the Performance CPU cores have 20% lower power consumption I doubt they are the same IP. TSMC's 4nm doesn't bring any efficiency over the N5P that the A15 uses. It brings minimal reduction in area and cost and that's it.
 

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
Pretty much what you'd expect if the A16 is the same core as A15 but with LPDDR5 instead of LPDDR4X and a 7% frequency boost (probably more due to better yield allowing for higher bin splits with Apple's "all or nothing" frequency cutoff binning)

I'll be curious to see how the little cores perform - are they are the same too or did Apple iterate them again like they did last year?