Discussion Apple Silicon SoC thread

Page 338 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
24,001
1,620
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:



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:


M4 Family discussion here:

 
Last edited:

Doug S

Diamond Member
Feb 8, 2020
3,133
5,380
136
Edit: No wonder Apple didn't mention any transistor counts during the event. A18 Pro's die size stayed about the same compared to A17 Pro on a lesser dense node (N3E vs N3B), so the transistor count might actually be less.

Finflex could allow them to have more transistors and less area even in a less dense process.
 
Reactions: Mopetar

SteinFG

Senior member
Dec 29, 2021
711
835
106
I expect the difference in size to diverge over time. Right now A18 Pro has 1 more GPU core, 1 more Prores encoder, USB 3.2 controller instead of USB 2.0, and like 20MB more cache (I think).
Next year A19 Pro could get fatter NPU/GPU/CPU; while leaving A19 to stall
 

Doug S

Diamond Member
Feb 8, 2020
3,133
5,380
136
Next year A19 Pro could get fatter NPU/GPU/CPU; while leaving A19 to stall

They're still limited by power, they can't keep adding cores - people don't want their Pro phones to get hot in their hands or have a shorter battery life than non-Pro. If they add cores the NPU seems like the most likely candidate, but they probably want to add resources to both Pro and non Pro if they want to push Apple Intelligence - same reason they went to 8 GB in the base iPhone. The NPU is tiny and only consumes power in short bursts so it is more amenable to beefing up than the GPU (which has already been done, twice, recently) or the CPU (which is more than competitive unless one cares about winning pointless MT contests on a phone)

One thing I'd expect them to diverge on is memory standards. They've been slow to adopt new memory standards on iPhone but this would give them an opportunity to advance new standards like LPDDR6 (say on A20P) while A20/A21 stick with LPDDR5X. They might do the same on the Mac, pushing faster memory controllers on Pro/Max/Ultra before they arrive on the base M.
 

Doug S

Diamond Member
Feb 8, 2020
3,133
5,380
136
look at the number of display engines hahah, still 2

Dunno why that's a problem. How many people want to connect more than one external display to an entry level laptop or more than two displays to a base Mini? Not disputing there aren't people who want to do that, but it is hardly a mainstream need and so making people upgrade from the base M to get it hardly terrible.

Since the PC world has always been chipset based instead of SoC based the capability wasn't part of the CPU, so a "low end" CPU was able to support more displays. But that was due to the way they split up the functionality, not because they were generous and thought there were a lot of people wanting 3 or 4 displays on a low end CPU. If PCs ever get rid of the "chipset" and that's subsumed into design as one of the chiplets, bet you see the same thing on Intel/AMD.
 

Doug S

Diamond Member
Feb 8, 2020
3,133
5,380
136
Wow I knew Apple was working on their own wifi and bluetooth but didn't think they were debuting those at the same time as their cellular modem. Plus GPS too.

Pretty ambitious. I'd give it about 24 hours after it goes on sale before the first breathless articles from "analysts" claiming Apple made a big mistake because some Youtube blogger did some tests and found speedtest shows Apple modem gets "only" 350 Mbps where a Qualcomm modem gets 500 Mbps, or they have fewer bars on wifi or they think the GPS fix is two meters less accurate

But I suppose you test this stuff on the low end model first for a reason!
 

jdubs03

Golden Member
Oct 1, 2013
1,235
871
136
Wow I knew Apple was working on their own wifi and bluetooth but didn't think they were debuting those at the same time as their cellular modem. Plus GPS too.

Pretty ambitious. I'd give it about 24 hours after it goes on sale before the first breathless articles from "analysts" claiming Apple made a big mistake because some Youtube blogger did some tests and found speedtest shows Apple modem gets "only" 350 Mbps where a Qualcomm modem gets 500 Mbps, or they have fewer bars on wifi or they think the GPS fix is two meters less accurate

But I suppose you test this stuff on the low end model first for a reason!
They’d have to be expecting that kind of negativity. I was always of the opinion that they wouldn’t release their own in-house designs until it’s at least competitive with its current component suppliers/competitors.
 

The Hardcard

Senior member
Oct 19, 2021
314
397
106

Interesting
That explains why a separate die instead of binning. A18 is going into everything as the base Apple Intelligence experience. iPhone SE, base iPad, iPad Mini, Apple TV, Homepod. Too much volume for the extra silicon not being used. Especially at sky high N3 prices.

The savings on that volume probably more than covers the costs of a separate die. It’s probably going to be an extended life SOC, probably on sale new in some device(s) in 2026.
 

jdubs03

Golden Member
Oct 1, 2013
1,235
871
136
That explains why a separate die instead of binning. A18 is going into everything as the base Apple Intelligence experience. iPhone SE, base iPad, iPad Mini, Apple TV, Homepod. Too much volume for the extra silicon not being used. Especially at sky high N3 prices.

The savings on that volume probably more than covers the costs of a separate die. It’s probably going to be an extended life SOC, probably on sale new in some device(s) in 2026
An 11th gen iPad with A18 becomes quite the value proposition. I’d probably buy one in place of my 9th gen pretty damn quickly. 5 generation improvement for the CPU plus min 8GB of RAM. No brainer. Something tells me it might be a tad pricier than the previous launches, but still, would be nice.
The 9th gen can still hang out as a secondary hah.

But yeah, Apple is definitely pushing into artificial intelligence with force so I agree with you that they’re trying to equip as much as they can with A18.

Idk about A18 in new devices 2 years from now though. Might upset some folks.
 

poke01

Diamond Member
Mar 8, 2022
3,441
4,718
106
I'll never buy Apple either, but I respect what they got out with their SoC. Lunarlake would never exist without Apple threat. I remember a magazine where it talked about Apple iPod batteries only lasting 18 months but couldn't be replaced.

One guy got pissed off so much that he went around all the Apple stores in his city and put up a sign "Apple iPod battery only last 18 months, and they don't give service nor allow you to replace them". Few months later they said they'll offer service. Not a coincidence I bet. You'll see Louis Rossman basically spearheading that all manufacturers embrace Right to Repair, and he's owns an Apple repair shop.

It's because I don't like the Apple ecosystem too much that I wished the x86 vendors would do much better.
Yeah Apple does pull crap like that. Honestly I would only bother with Apple products where comsumer law is strong, so EU and Australia/NZ. Apple will have to fix the problem and cannot fully force customers into paying $$$$ but provide any fix for free based on the cost of product. They can be excellent products if they make them easier to repair/upgrade.

We come a long way since those iPod days, time files. The new iPhone 16 battery method is proof of that, the right to repair laws are working. I do hope more countries follow suit.

 
Last edited:

name99

Senior member
Sep 11, 2010
602
491
136
There is no information, but there is a test from Geekerwan.
You can try to compare these with https://chipsandcheese.com/2024/09/19/running-spec-cpu2017-at-chips-and-cheese/
Which after some scrolling gives values for the latest AMD and Intel.
Essentially they are the same pattern though, for the hardest cases (which are the ones that matter most!), the Apple number is slightly better.

But I’m not sure I trust that the two are exactly comparable. Branch prediction now has multiple stages and I’m not sure that people running these tests are collecting everything you’d want to know.
eg, for Apple, the prediction can fail at the Fetch stage, then be corrected a cycle later by a simple direction predictor, then recorrected three cycles later by a TAGE predictor.
Do we score as better an overall “eventually the right value was predicted” or some sort of weighted value of how often we lost a cycle or two in missed predictions before we re-steered?

I think this is an area where we need academic investigation by people who consider all the issues, not just reading what the (differently defined by different companies) performance monitoring registers give.
 

Eug

Lifer
Mar 11, 2000
24,001
1,620
126
Dunno why that's a problem. How many people want to connect more than one external display to an entry level laptop or more than two displays to a base Mini? Not disputing there aren't people who want to do that, but it is hardly a mainstream need and so making people upgrade from the base M to get it hardly terrible.

Since the PC world has always been chipset based instead of SoC based the capability wasn't part of the CPU, so a "low end" CPU was able to support more displays. But that was due to the way they split up the functionality, not because they were generous and thought there were a lot of people wanting 3 or 4 displays on a low end CPU. If PCs ever get rid of the "chipset" and that's subsumed into design as one of the chiplets, bet you see the same thing on Intel/AMD.
It seems HDMI on the base M3 is still limited, although I’m not entirely sure since the Mac mini is still M2. If the M4 removes the HDMI limitations of the M2 Mac mini, that would very welcome.

M3 MacBook Pro specs:

Display Support
M3
Simultaneously supports full native resolution on the built-in display at 1 billion colours and:
One external display with up to 6K resolution at 60Hz


M2 Mac mini specs:

Display Support
Simultaneously supports up to two displays:
One display with up to 6K resolution at 60Hz over Thunderbolt and one display with up to 5K resolution at 60Hz over Thunderbolt or 4K resolution at 60Hz over HDMI
Thunderbolt 4 digital video output
Support for native DisplayPort output over USB‑C
HDMI display video output
Support for one display with up to 4K resolution at 60Hz
 
Reactions: ikjadoon

DZero

Golden Member
Jun 20, 2024
1,034
383
96
Wow I knew Apple was working on their own wifi and bluetooth but didn't think they were debuting those at the same time as their cellular modem. Plus GPS too.

Pretty ambitious. I'd give it about 24 hours after it goes on sale before the first breathless articles from "analysts" claiming Apple made a big mistake because some Youtube blogger did some tests and found speedtest shows Apple modem gets "only" 350 Mbps where a Qualcomm modem gets 500 Mbps, or they have fewer bars on wifi or they think the GPS fix is two meters less accurate

But I suppose you test this stuff on the low end model first for a reason!
Imagine if somehow Apple manages to tie with Mediatek, which are below Qualcomm but way above Exynos.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |