Discussion Apple Silicon SoC thread

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:

Panino Manino

Senior member
Jan 28, 2017
813
1,010
136
Did they say 128KiB+192KiB L1 cache?
That's huge, Zen and Skylake are 32KiB+32KiB L1.

This huge amount of cache is intended to translate x86 code?
At least it can help, right?

Anyway, it's over.
The lock is in place and closed, now Macs are a fully closed system. Nothing is expandable and nothing can be run without Apple's blessing.
 
Last edited:

name99

Senior member
Sep 11, 2010
404
303
136
M1
5 nm
Unified memory architecture
16 billion transistors
192 KB instruction cache
128 KB data cache
Shared 12 MB L2 cache
8-core CPU - 4 high-performance, 4 high-efficiency
(The 4 high-effiency cores perform like dual-core Intel MacBook Air)
Integrated GPU - 8-core 2.6 Teraflops
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

And, apparently, LP-DDR4.
That seems strange. For iPhone I wasn't surprised because the volumes necessary may not be available. But for MBA the volumes should be available.
Is it proving disappointing in other usage cases? (ie not much speed boost, or surprisingly high energy usage)?
 

trivik12

Senior member
Jan 26, 2006
250
201
116
even ram seem to be integrated into SOC !!! Also its LPDDR4x and not LPDDR5. But yearly updates like iphone. I guess we will see new macs every year like iphone.

Looking forward to Anandtech review on all these products. I hope Apple sends them review units. I hope its as fast as Apple claims. I hope we can run exactly the same benchmarks as we do in x86.
 

name99

Senior member
Sep 11, 2010
404
303
136
Probably only for the MBA and the 13" MBP.

And mac mini. But everything about that mac mini (fewer USB ports, 4 cores, only 1G ethernet) means they'll also ship a mac mini pro.

I think the only question now is do we get 8 core machines (iMac, mini pro) based on a bumped up M1, or do those wait till next year and the A15 based SoCs.
iMac in particular I can't believe will wait a year. There has to be another event left.
This M1 seems basically a rebranded A14X. There's still the mythical A14T and Lifuka left; my guess is we see them ? February ? April ?

Presumably in terms of PR Apple want to maximize things. Let these drop, get two months of "Damn, Apple is the king -- but if you want throughput you still want those 16 core AMD chips" followed by the 8 core machines and an Apple dGPU?

(And we still need an event for new aTV and iPad Pros, god damnit! Those are what I want to buy right now!!!)
 

tamz_msc

Diamond Member
Jan 5, 2017
3,697
3,547
136
M1 iGPU is mobile GTX 1050Ti level with those Tflops and Pixel and Texel fill rates. Pretty impressive, though I noticed that Baldur's Gate 3 was running at a lower quality level and there were frame rate issues with some of the games they showed.
 

name99

Senior member
Sep 11, 2010
404
303
136
even ram seem to be integrated into SOC !!! Also its LPDDR4x and not LPDDR5. But yearly updates like iphone. I guess we will see new macs every year like iphone.

Looking forward to Anandtech review on all these products. I hope Apple sends them review units. I hope its as fast as Apple claims. I hope we can run exactly the same benchmarks as we do in x86.

That was not surprising. If you want the performance of LPDDR, the really high frequencies, you have to do that, it's just physics. This is the future of computing.
(Just like in the old days, say mid 90s, your L2 was an external SRAM chip and you could change its size. But you give up that flexibility for a massively faster on-chip L2.)
 
  • Like
Reactions: ikjadoon

name99

Senior member
Sep 11, 2010
404
303
136
Unlikely. We don't get that for iPads for example.


That's a disappointment for the MacBook Pro. Makes sense for the fanless MacBook Air.

And as expected, no 12" MacBook. :(

These are very much the three low-end machines. Low-end means, uh, low-end...

Don't waste time complaining that high-end doesn't exist; the only interesting question is how many months until Apple announce it.
 

jpiniero

Lifer
Oct 1, 2010
14,487
5,150
136
I think the only question now is do we get 8 core machines (iMac, mini pro) based on a bumped up M1, or do those wait till next year and the A15 based SoCs.
iMac in particular I can't believe will wait a year. There has to be another event left.
This M1 seems basically a rebranded A14X. There's still the mythical A14T and Lifuka left; my guess is we see them ? February ? April ?

Current iMac was refreshed in August so they have time.
 

IvanKaramazov

Member
Jun 29, 2020
56
102
66
The MBA replaces all Intel models, but the MBP only replaces the lower-end, two-port Intel models, and the high-end Intel models remain in the lineup. Same for Mac Mini.

The M1 is probably an A14X with added IO and such, and I'm guessing the 16GB RAM ceiling and the 2 port limit is associated with that. I expect the M1 will be faster for both CPU and GPU than Tiger Lake, but it's pretty clearly a first foray into the world of Arm Macs. Will be interesting to see if the Bloomberg 8x4 core, which will certainly appear in the 24" iMac next year, also gets backloaded into the high-end Macbook Pro 13 and Mac Mini models. Otherwise we're probably waiting for Fall 2021 to see 13" Macbook Pros with 32 GB RAM, 4 ports, etc.

Lots of big numbers thrown around r.e. performance and battery life. I'm sure they'll be impressive, but it's all essentially handwaving until these things get into the hands of reviewers.
 

name99

Senior member
Sep 11, 2010
404
303
136
M1 iGPU is mobile GTX 1050Ti level with those Tflops and Pixel and Texel fill rates. Pretty impressive, though I noticed that Baldur's Gate 3 was running at a lower quality level and there were frame rate issues with some of the games they showed.

Were the games they showed running native? Or were they x86 originals deliberately shown to make the point that "we're good enough, even at emulation"?
You get my point; emulated doesn't show us what the GPU can do under ideal circumstances.