Discussion Apple Silicon SoC thread

Page 336 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
23,807
1,385
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:


M4 Family discussion here:

 
Last edited:

moinmoin

Diamond Member
Jun 1, 2017
5,064
8,032
136
Part of the reason efficiency cores are much more efficient is that they have a more limited instruction set, which is less of a problem for someone like Apple since the control both hardware and software and can design both with a shared goal in mind.
The biggest reason Apple doesn't have this problem is that there is no deviation between the two kinds of cores. What Intel rather arbitrarily packs into the x86 instruction set (and AMD so far sadly doesn't dare to deviate from), Apple packs into separate dedicated and as such way more efficient accelerators and engines on the same SoC.
 

eek2121

Diamond Member
Aug 2, 2005
3,100
4,398
136
Interestingly they state that CPU TDP is 30W and 55W for the GPU.

Claimed efficency seems to me a bit optimistic, we ll see how it does in real benches, though.

That GPU perf vs. power graph is complete marketing nonsense. Read the fine print, they didn’t compare it with the 3070, they compared it with Intel’s Integrated graphics! Talk about lipstick on a pig.

EDIT:

1634593194975.jpeg
 

BorisTheBlade82

Senior member
May 1, 2020
680
1,069
136
As long as Apple has access to TSMCs next gen node, and don't do a bad job of designing a cpu and gpu, they're going to be more efficient than their pc counter parts. If we had an AMD 5 nm based laptop cpu and a 5 nm GPU, the numbers would be different. So kudos to Apple for pushing new technology. Personally I just want to buy a video card at MSRP :p
That is simply not true. At best about 25% of their advantage is caused by the process. Just take a look at what TSMC officially claims and then take a look at Apple's superiority again.
 
  • Like
Reactions: Schmide

Hitman928

Diamond Member
Apr 15, 2012
6,122
10,523
136
It may not be possible to get any better reviews. How much "real world" software is there in common between MacOS and Windows or Linux? It's almost not worth complaining about it anymore, though. Some people just want their SPEC scores.

Meanwhile:


(check the M1 Max score for CBR23)

vs:


(check the CBR23 MT score for the 5980HS)

But:

117495.png


So in 526.blender_r and 511.povray_r, the M1 Max massacres the 5980HS, but in Cinebench R23 it . . . doesn't? Wouldn't it be nice of AT to run a current build of Blender on their M1 Max review sample instead?

SPEC usually does custom compilations of each software it benchmarks and tends to compile more for the least common denominators. For instance, in the x264/x265 tests they compile without AVX or SSE flags on x86 which is not at all how the software is actually used. This could paint a very different picture of video encoding performance as x86 kind of relies on those more modern instructions. Unless Anandtech is custom compiling the sub tests and turning these flags back on? Either way, some of the tests make it hard to really compare performance for real world apps when they're not testing the apps in the way they are actually compiled/run in the real world.
 

eek2121

Diamond Member
Aug 2, 2005
3,100
4,398
136
There’s a lot of people who hate to admit Apple has made the gains they have. Heck, ive hated Apple computers for years myself. Too expensive, lack of upgradability, closed eco system, etc. There’s plenty of reasons to dislike their practices. They’ve finally made a suitable product for a bigger market share than they’ve had and it’s tough to get on their side. I’m still reading innuendo that is focused completely on disproving their gains and looking for the “I got ya” piece of data to say “I told ya so”. I get it but even begrudgingly one had to admit the obvious. Cheers…

I hope you aren't referring to me. Quite frankly, I'm glad they were able to pull off what they did, but too many people here overhyped it and pretends like it's god's answer to everything that is "wrong" with the PC market. There are users getting sucked into this that realize, far too late, that their new shiny Macbook can't do what their old PC did. I'm one of the most objective and unbiased people you will meet when it comes to hardware. I've said both good and bad things about every major company.

If Apple would allow an ARM port of Windows on the Mac and provide GPU drivers for Windows, they'd get a lot more support from many of us.

The concerns I've attempted to state previously are valid:
  1. Limited software support. (macOS only, some proprietary macOS software doesn't work under Rosetta as well, no native Linux dual boot support except third party without driver support)
  2. Overpriced for what you get. (A PC equivalent would cost $1,000 less and run WAY more software, and depending on the software, can run that software faster. I can provide links if needed. Amazon has lots of Ryzen laptops with GeForce graphics)
  3. Performance gains drastically overstated by certain members of this forum repeatedly. (GPU is "around" 3060 level NOT 3080/3090, CPU is NOT 500% more efficient than AMD/Intel, or even 3x-4x like some people here are claiming. Given that Apple is on a newer node and just released a brand new product on that node, being ahead is expected. Zen 4 will easily close the gap. Alder Lake Mobile appears to already beat the M1 Pro Max in terms of raw performance in early benchmarks, and it's on an older node. Translation: I would hope Apple can be competitive with year old technology)
 

Heartbreaker

Diamond Member
Apr 3, 2006
4,334
5,451
136
18% is good, given the process limitations Apple was facing (and they are probably still recovering a bit form the brain drain that hit their semiconductor group. Folks are just being salty.
I did want to see info on the M2 Pro today. That would really make the Mac Mini a viable lower cost option to the Mac Studio.

18% is multi-core though. Much like A15, the single core is likely too negligible to mention. Gains are probably more from the efficiency cores.
 

Doug S

Platinum Member
Feb 8, 2020
2,742
4,664
136
But that might just be skipping M2 in the midrange and above. So could be about 2 years since the last upper range chip.

The point here is cadence of tapeouts for the product ranges.

You can do an entry chip every ~18month average (or even shorter) since it's also the iPad chip. This is the highest volume M chips, so that's no problem at all.

But once you hit midrange, volume drops a lot, so cadence likely will as well. I'd expect a 2 year cadence average or longer.


They can't skip M2 for the high end, that's what the Mac Pro will be based on. The M1 Max die is able to link to only one other M1 Max to create the M1 Ultra. They need an M2 Max with three off chip I/O pads to create the "M2 Extreme" that will go in the Mac Pro. They can't build it out of M1.

Given that M1 had only two mask sets - the M1 and the M1 Pro/Max which was the same die just "chopped" when making M1 Pros I don't see any reason why they can't have the full line for each generation. If they were doing Mx yearly I could see them only doing the the Pro/Max every other year, but this has been more like 18 months as the rumors claimed. If they maintain that there's no way they go three years between new Pro/Max/Ultra/Extreme offerings.
 

MadRat

Lifer
Oct 14, 1999
11,942
264
126
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
Jul 27, 2020
19,823
13,588
146
Since entry level is only 8GB RAM too, another consideration is how swapfile usage eats into the endurance.
Good point. Heavy usage would have a double whammy effect on the life of the SSD.

AFAIK, aren't Apple being characteristically silent and don't actually state the endurance of their soldered SSDs?
That's coz most of their sheep, errr...users :D, would just send in the laptop for repairs rather than put two and two together and figure out that Apple made a nice buck by making them pay a lot for a puny SSD with a shorter lifespan. In the x86 world, users are very cost conscious and vocal about transparency so every decent company in the x86 ecosystem tries to be as forthright as possible to stay in the good books of techies since they are the ones recommending products to their clients/friends/families.
 
  • Like
Reactions: Lodix and KompuKare

poke01

Platinum Member
Mar 8, 2022
2,094
2,626
106
Since entry level is only 8GB RAM too, another consideration is how swapfile usage eats into the endurance.
AFAIK, aren't Apple being characteristically silent and don't actually state the endurance of their soldered SSDs?
Apple uses SK hynix for the Macbook Air and Kioxia for the MacBook Pro(Apple Sillcon models)
Source: iFixit Chip ID webpage

As for drive indurance that's up to user to know using smartmontools using the macOS CLI.
For reference heres my 16" Intel MacBook Pro 2019 SSD info that I been using since 2020 January, almost everyday till now.
I have a 512 SSD and its made by Toshiba.

It's been 2 years and 7 months and I used 1% of my SSD. I have written 32.1TB so far.
I would say the endurance is very good.

1659873192211.png
1659873071913.png
 

Doug S

Platinum Member
Feb 8, 2020
2,742
4,664
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

Mopetar

Diamond Member
Jan 31, 2011
8,104
6,730
136
Uh, no. The last few cores are designed from the Haifa, Israel team.

Also the people responsible for amazing chips in Apple have left. That's big reason why the last few chips have shown less than expected gains.

So what do those people do? They go work at other companies. It's not always at direct competitors, and it could be a startup. But in lots of cases they do switch to direct competitors and it literally makes the others better.

These seasoned veterans of engineers are very limited in numbers so by moving from one company to other you could say one company is "sucking" the lifeforce out of other to become better.

Apple is one of the richest companies on the planet. If they really felt that there were a few dozen people who would let them make a chip significantly better than they otherwise can, they can pick up the phone and get those people on the job within weeks.

If that's all it was, Apple would have done it. Pay 100 engineers another $250,000 each every year and you're only spending an extra $25 million. Chump change for them.
 

Henry swagger

Senior member
Feb 9, 2022
496
301
106
Last edited by a moderator:

FlameTail

Diamond Member
Dec 15, 2021
3,855
2,297
106
Apple Silicon is frankly incredible.

If we compare the Apple M2 to a 6-core Ryzen such as R5 6600H (this is reasonable because M2 is 4P+4E, while a 6600H has 6P) in Geekbench (the best benchmark for this kind of comparison):

M2 - 1900/9000

6600H - 1500/8000

The M2 outperforms the 6600H, particularly by a large margin in single core, while consuming around half the power the 6600H does.

This is crazy, and can't all be chalked down to the M2's superior process node...The M2 certainly has a significant architectural advantage.
 

moinmoin

Diamond Member
Jun 1, 2017
5,064
8,032
136
Something to keep in mind (and most here keep hinting at already) is that nodes at TSMC are not as clear cut as it's made seem. What TSMC essentially advertises publicly is for what TSMC seeks additional customers. Existing big customers like Apple, MediaTek and AMD etc. that command big volumes work closer with TSMC anyway so the nodes they use can become more of a mix of everything and possibly include changes that TSMC only later makes public as part of a newer node. Strict comparison of node timelines and features with such bigger customers become patchy as a result.
 

Glo.

Diamond Member
Apr 25, 2015
5,802
4,776
136
I would bet the non M3 remains 128 bit if M3 Pro goes to 384. That goes into stuff like iPad Pro & Macbook Air that are passively cooled.
IMO, M3 is either 192 or... 256 bit.

You need all of that bandwidth for higher CPU thoughput, and especially for RT capabilities in the GPUs. Macroscalar architectures have one problem. Heavy demands for memory bandwidth.
 
  • Like
Reactions: A///

Tigerick

Senior member
Apr 1, 2022
695
601
106
I’m not confident in Mark Gurman’s prediction this time around. I could very well be wrong, but I’m not convinced they will be using N3/N3B for M3.
Yeah, release timings and yield might be issues, that's why I omit M3-based MacBook Air 13 & 15. If M3 is based on N3E, we might be expecting first release around one year later (might be earlier depends on yield), ie end of 2024.

That is the reason I think Apple might be focusing on iMac and MBP13 which are not high-volume Macs. I have created a thread @ Macrumor Forum to discuss potential higher price point on upcoming MacBook Pro 13 here. You are welcomed to pitch in :p
 
Last edited:

Doug S

Platinum Member
Feb 8, 2020
2,742
4,664
136
This is completely wrong and ignores countless historical examples of countries that have tried to get out from under their debt by running the printing press. Both Zimbabwe and Venezuela are recent examples of this. Their governments kept creating money at rates vastly outpacing their own economic growth, which caused further damage to their own economies making the problem worse.

If you don't understand the massive difference between increasing the money supply and actually printing physical cash in larger and larger denominations, there's no point in continuing this discussion.
 

repoman27

Senior member
Dec 17, 2018
381
536
136
They've been doing three separate designs since M1.

M3 Pro got gutted.

TSM has been good at ramping nodes into volume since N7.
Not really an achievement?
Lol. As usual, you're confidently wrong about everything, because you fundamentally lack knowledge of the space. I've seen your post history so I'm not even going to bother to engage.

Which part of the M3 is a trainwreck meme again? The part where Apple hard launched products containing a family of three SoCs manufactured using first generation N3 just 10 months after the process reached volume production and right on time for the most important sales quarter of the year?

How does that compare to Intel completely missing 2023 including the holiday quarter with MTL, which is mostly fabbed by TSMC and mostly on legacy nodes anyway? How does it compare to Qualcomm paper launching the TSMC N4 Snapdragon X Elite more than 6 months out from release?

Don't get me wrong, I'm all for competition and what that brings, but Apple and TSMC just threw down a gauntlet here and Intel and Qualcomm need to start executing.
 
  • Like
Reactions: Tigerick and Viknet

FlameTail

Diamond Member
Dec 15, 2021
3,855
2,297
106
AFAIK, LPDDR5 dies are available in 8, 12, 16, 24, and 32 Gbit densities. 18 GB packages are totally a thing, but I believe they are all made using 6x 24 Gbit dies. As I said, I thought it unlikely that Apple would go to 6-high stacks.

chrome_screenshot_1698943127793.png
 

Glo.

Diamond Member
Apr 25, 2015
5,802
4,776
136
Hmm, he didn't make any subjective comments about it being blurry or smearing. Have you used one in person?
Yes, I have used one in person. Without it I would not be commenting. I used one BEFORE I found out the response times, tho. Which explained what I have seen.
 
  • Like
Reactions: jt7 and Ajay

FlameTail

Diamond Member
Dec 15, 2021
3,855
2,297
106
How about this:

Apple makes Base M chip every year, whereas M Pro, M Max, M Ultra are released every two years.

This makes sense because base M chip goes into a lot of high volume consumer prodcucts: iPads, Macbook Airs, iMacs, Mac Minis and Vision headsets.

Would this strategy make sense?
 

Eug

Lifer
Mar 11, 2000
23,807
1,385
126
But 18 months makes no sense for a product cycle for Macs, since there are specific timings the market prefers like "back to school" and "Christmas". There's a very good reason Apple introduces new iPhones at the same time every year, and those reasons apply to other products. Something like a Mac Pro is probably largely independent of that due to who is buying it, but for the lower end (sub $2000 in Apple world) of the product lines where the large majority of units will be sold their marketing unit would likely overwhelmingly desire the same release time every year. Meaning either 12 or 24 months, not 18.
Average release interval of the iMac is 17 months.
Average release interval of the Mac mini is 23 months.
Average release interval of the 13” MacBook Air is 15 months.
Average release interval of the regular iPad is 20 months.
Average release interval of the iPad Air is also 20 months.

As for the Mac Pro, it’s 31 months.
 
  • Like
Reactions: Tlh97 and dr1337