Discussion Apple Silicon SoC thread

Page 140 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
23,579
992
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:

poke01

Senior member
Mar 8, 2022
633
603
106
Does the 20B transistors in the M2 also include the Nueral Engine?
Because the M2 has a beefy Nueral Engine compared to 6800U and 1260P.

Since the 6800U is chiplet design does it include the I/O as part of the transistor count?

Also what is the power saving between 5NP and N6?
 

repoman27

Senior member
Dec 17, 2018
342
488
136
Does the 20B transistors in the M2 also include the Nueral Engine?
Because the M2 has a beefy Nueral Engine compared to 6800U and 1260P.

Since the 6800U is chiplet design does it include the I/O as part of the transistor count?

Also what is the power saving between 5NP and N6?
The 20B for the M2 is all-in, the whole SoC.

The 6800U is actually a monolithic SoC as well, not chiplet, and includes I/O:
Rembrandt-APU.jpg


I believe N5P is roughly 33% lower power at the same performance as N6, but TSMC was pretty vague about power and performance for N6.
 

repoman27

Senior member
Dec 17, 2018
342
488
136
Also, the proper Intel comparison for the M2 is something based on the Alder Lake 2+8 die, like the Core i7-1265U. The M2 can scale down to work in a passively cooled 10.9-inch iPad. The TDP for the ADL 6+8 die doesn't go lower than a minimum assured power (cTDP down) of 20 W.
 

Doug S

Platinum Member
Feb 8, 2020
2,191
3,380
136
So Apple uses roughly 175% of the silicon compared to AMD, and about 190% compared to Intel, per CPU? Seems like AMD and Intel are also on inferior manufacturing processes, so there is that. I'm less impressed by the performance metric results now.

Now do the math again, counting only the units that the AMD and Intel chips you are comparing to have. If those have no GPUs, you can scratch that off (and the DP/HDMI area) off the Apple Silicon die area. Also scratch off the NPU, IPU, and probably some other blocks too.

Now that comparison doesn't look anything like your claim, does it?
 

MadRat

Lifer
Oct 14, 1999
11,908
228
106
The raw size of the silicon on the chip already clues you in that Apple counts transistors differently between marketing and engineering. They probably are counting only the fraction of the wafer that directly compares with the competitor's solution. And whenever possible the marketing alludes to base products found in cell phones when marketing Mac Books. I noticed marketing by Apple seems to push the M1 cores with ambiguous meaning, giving a picture that all M1s are the same. The M2 is being marketed the same. But looking at more specific products, some versions are several times the size of the base product. So when you hear an M1 is x number of transistors they could be talking an A15 when the product may in fact use a much bigger form of the M1. Very misleading. Honestly, the M2 smells like a die shrunk A15 with a pinch of creative marketing.
 
  • Haha
Reactions: lobz

repoman27

Senior member
Dec 17, 2018
342
488
136
<utter nonsense>
The M series of SoCs uses several different dies. So far Apple has released:

M1 (Tonga)
M1 Pro (Jade Chop)
M1 Max (Jade)
M2 (Staten)

Apple-M1-chip-family-lineup-220308.jpg

Apple-WWDC22-M2-chip-M1-chip-2up-220606.jpg
Note: the second image is not at the same scale as the first.

None of them are used in iPhones. That would be the A series. The A14 shares the same microarchitecture and manufacturing process as the M1 family, likewise the A15 uses the same cores and manufacturing node as the M2. The M series chips are larger than the corresponding A series. The transistor counts from Apple represent the whole die. Nobody is counting anything differently. The M1 Ultra is made from two M1 Max dies joined via a local silicon interposer bridge.

The transistor counts from Apple are:

A14 Bionic: 11.8B
M1: 16B
M1 Pro: 33.7B
M1 Max: 57B
M1 Ultra: 114B
A15 Bionic: 15B
M2: 20B

The die sizes according to measurements by TechInsights and my own calculations are:

A14 Bionic: 88 mm²
M1: 119 mm²
M1 Pro: 230 mm²
M1 Max: 390 mm²
M1 Ultra: 780 mm²
A15 Bionic: 108 mm²
M2: 148 mm²
 
Last edited:

poke01

Senior member
Mar 8, 2022
633
603
106
The raw size of the silicon on the chip already clues you in that Apple counts transistors differently between marketing and engineering. They probably are counting only the fraction of the wafer that directly compares with the competitor's solution. And whenever possible the marketing alludes to base products found in cell phones when marketing Mac Books. I noticed marketing by Apple seems to push the M1 cores with ambiguous meaning, giving a picture that all M1s are the same. The M2 is being marketed the same. But looking at more specific products, some versions are several times the size of the base product. So when you hear an M1 is x number of transistors they could be talking an A15 when the product may in fact use a much bigger form of the M1. Very misleading. Honestly, the M2 smells like a die shrunk A15 with a pinch of creative marketing.
No you are mistaken, Apple is right in their transistor count when marketing.

A15 = 15 Billion transistors
M2 = 20 Billion transistors

A14 = 11.8 Billion
M1: 16 billion;
M1 Pro: 33.7 billion;
M1 Max: 57 billion;
M1 Ultra: 114 billion.

The M2 is an A15 base chip. It adds 2 more P cores and 2 extra GPU cores and better memory controllers when compared to A15.
 

Eug

Lifer
Mar 11, 2000
23,579
992
126
Meanwhile I'm still wondering when we can actually buy an M2 Mac mini.

I wonder if it's due to an SoC constraint issue, other chip constraint issue, or marketing choices. I suspect it may be all of the above.

An M2 Pro release soon would be nice too.
 

Heartbreaker

Diamond Member
Apr 3, 2006
4,219
5,220
136
Meanwhile I'm still wondering when we can actually buy an M2 Mac mini.

I wonder if it's due to an SoC constraint issue, other chip constraint issue, or marketing choices. I suspect it may be all of the above.

An M2 Pro release soon would be nice too.

Mac Mini neglect is just business as usual for Apple. Most people seem to blame Intel for this in the past, but I always thought Mini was an afterthought for Apple.
 

Doug S

Platinum Member
Feb 8, 2020
2,191
3,380
136
Mac Mini neglect is just business as usual for Apple. Most people seem to blame Intel for this in the past, but I always thought Mini was an afterthought for Apple.

Apple considers laptops their primary market, so it isn't a surprise that a newly released chip comes out on laptops first.

The surprise was the M1 coming out on the Mini out of the gate, but that may have been a nod to developers as a replacement for the A12Z based developer hardware that came in a Mini form factor.
 

richardskrad

Member
Jun 28, 2022
51
44
51
Apple is an 800 pound gorilla but they have been moving like a start-up ever since they took their destiny in their own hands and moved on from Intel. Tim Cook is not a product guy like Steve Jobs was but Steve Jobs was no match for Tim Cook when it comes to being a CEO.
 
Last edited:

MadRat

Lifer
Oct 14, 1999
11,908
228
106
From the YouTube reviews, unless you have the chassis in the larger MacBooks the M2's seems to be throttle prone. The Mini might have similar concerns.
 

repoman27

Senior member
Dec 17, 2018
342
488
136
From the YouTube reviews, unless you have the chassis in the larger MacBooks the M2's seems to be throttle prone. The Mini might have similar concerns.
Once again, the microarchitectures in the M2 are more power efficient than those found in the M1—they offer higher performance while using less power. The iPhones 13 are consequently less prone to throttling than the previous generation. The M2 is also designed to be used in a passively cooled 10.9-inch iPad Air with strict limits on skin temperature. The current Mac mini chassis can accommodate an Intel Core i7-8700B with a 65 W TDP and is capable of dissipating at least 122 W. Do you really think a ~9 W TDP M2 is going to have issues in that chassis?
 

poke01

Senior member
Mar 8, 2022
633
603
106
Once again, the microarchitectures in the M2 are more power efficient than those found in the M1—they offer higher performance while using less power. The iPhones 13 are consequently less prone to throttling than the previous generation. The M2 is also designed to be used in a passively cooled 10.9-inch iPad Air with strict limits on skin temperature. The current Mac mini chassis can accommodate an Intel Core i7-8700B with a 65 W TDP and is capable of dissipating at least 122 W. Do you really think a ~9 W TDP M2 is going to have issues in that chassis?
The M2 can go up to 30Watts in macbooks. It is no means a 9 watts SoC.
The M1 in the iPad Pro can go up to 20 watts while playing 3d games again higher the 9TDP.

For passive cooling it should stop at 10watts. Apple likes to cheap out on cooling in its macs. Making a SoC that can go up to 30watts with no cooling in the MBA or rather poor cooling in the 13" MBP is proof of this.

The previous Apple intel macs also throttled cause of poor cooling. Apple is not new to this
 

Doug S

Platinum Member
Feb 8, 2020
2,191
3,380
136
The M2 can go up to 30Watts in macbooks. It is no means a 9 watts SoC.
The M1 in the iPad Pro can go up to 20 watts while playing 3d games again higher the 9TDP.

For passive cooling it should stop at 10watts. Apple likes to cheap out on cooling in its macs. Making a SoC that can go up to 30watts with no cooling in the MBA or rather poor cooling in the 13" MBP is proof of this.

The previous Apple intel macs also throttled cause of poor cooling. Apple is not new to this


The maximum it can go is not comparable to Intel's TDP figures in any way. It is compared to their maximum power which is generally more than double their rated TDP.
 
  • Like
Reactions: MadRat

poke01

Senior member
Mar 8, 2022
633
603
106
The maximum it can go is not comparable to Intel's TDP figures in any way. It is compared to their maximum power which is generally more than double their rated TDP.
I know. When under load the M2 can use up to 30watts. All I am saying is the Air could use better cooling say a vapour chamber for a fanless design.
 

ZGR

Platinum Member
Oct 26, 2012
2,052
656
136
Apple has a low power mode already implemented in MacOS/iOS fortunately. It appears to clock the CPU down to 2 GHz. So if I was experiencing high temperatures, I'd enable that.
 

gdansk

Golden Member
Feb 8, 2011
1,944
2,281
136
Regarding the thermal throttling: It's much ado about almost nothing.
It's more a problem that people are using the MBA to replace the MBP13. The MBA is trying to be is the best thin and quiet laptop. If you want better sustained performance, buy something designed for that. In the Mac Mini they'll have a fan, so it won't be a problem.

For the Mac Mini I suspect the delay is either:
1. They designed a new, smaller chassis (wishful thinking?)
2. They expect to sell a lot of MBA and so delay the Mac Mini update until that demand reduces (in fall)
 
Last edited:

lobz

Platinum Member
Feb 10, 2017
2,057
2,856
136
Regarding the thermal throttling: It's much ado about almost nothing.
It's more a problem that people are using the MBA to replace the MBP13. The MBA is trying to be is the best thin and quiet laptop. If you want better sustained performance, buy something designed for that. In the Mac Mini they'll have a fan, so it won't be a problem.

For the Mac Mini I suspect the delay is either:
1. They designed a new, smaller chassis (wishful thinking?)
2. They expect to sell a lot of MBA and so delay the Mac Mini update until that demand reduces (in fall)
Exactly! If you want sustained performance, buy something even more(!) shamelessly overpriced! I think we all got the message loud and clear.
 
  • Haha
Reactions: igor_kavinski

gdansk

Golden Member
Feb 8, 2011
1,944
2,281
136
Exactly! If you want sustained performance, buy something even more(!) shamelessly overpriced! I think we all got the message loud and clear.
Overpriced? Other than some near-mythical 6800U laptops nothing is close. Blame their lack of competition. Apple is the only one making chips that aren't garbage when at 10W
 

ZGR

Platinum Member
Oct 26, 2012
2,052
656
136
Regarding the thermal throttling: It's much ado about almost nothing.
It's more a problem that people are using the MBA to replace the MBP13. The MBA is trying to be is the best thin and quiet laptop. If you want better sustained performance, buy something designed for that. In the Mac Mini they'll have a fan, so it won't be a problem.

"Just avoid holding it in that way." LOL, sorry couldn't resist.
 

gdansk

Golden Member
Feb 8, 2011
1,944
2,281
136
"Just avoid holding it in that way." LOL, sorry couldn't resist.
If you think those Alder Lake laptops, with their 64W PL2, are any better then prepare for a shock. When they drop to PL1 they lose more performance than this. And they're hardly even faster before throttling. With a fan (or two).
Has anyone even managed to ship a fanless Alder Lake laptop? Let's see how that SoC competes.
 
Last edited:

ZGR

Platinum Member
Oct 26, 2012
2,052
656
136
If you think those Alder Lake laptops, with their 64W PL2, are any better then prepare for a shock. When they drop to PL1 they lose more performance than this. And they're hardly even faster before throttling.
Has anyone even managed to ship a fanless 12x0U Alder Lake laptop? Let's see how that SoC competes.

Oh I wasn't trying to say it is worse than Intel. My experience with Intel based laptops has been quite a journey, but I do love tweaking them. I just don't like the idea that you shouldn't run compute heavy tasks on a MBA. You can, it just requires the user to know when they should manually throttle their system. I love that Steve Jobs quote.

Without 3rd party software, my Intel based Macbook Pro with its 45w i7 and 45w dGPU running an older version of MacOS performs terribly. I need to run MacsFanControl for the custom fan profile, Turbo Boost Switcher to disable Intel Turbo Boost, and gfxCardStatus to force the use of the iGPU or dGPU when I need to. Intel Turbo Boost loves to run at max turbo until it throttles. Once it hits 100C, it downclocks to 1.2 GHz, before hitting max turbo again. This creates a stuttering effect in performance which is extremely noticeable in games.

MacOS and Windows provide the ability to throttle the CPU. In Windows, we can force a CPU down to 550-800 MHz using 3rd party software, or disable boost in power plan. It'd be curious to see how much power an Alder Lake 12U laptop consumes at such a low frequency. This fixes throttling, but the average user won't be doing that easily.

MacOS' UI for throttling the CPU is far better compared to the ancient Windows Power Plan UI. But once I tweak an Intel laptop, throttling will go away. Most recent laptop I work with is an i9-9750H paired with a GTX 1650 ti. The "i9" will clock over 4 GHz quite often under heavy load, reaching 100C very fast. This causes the 1650 ti to also reach throttling temps. But once turbo boost is off, I never see it reach 80C and the 1650 ti is locked at max boost. I am amazed that I have been dealing with this overly aggressive Turbo Boost behavior all the way back to my 2009 laptop with its i7-720QM and GT 240M combo.
 
  • Like
Reactions: igor_kavinski