Discussion Apple Silicon SoC thread

Page 243 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
23,587
1,000
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:

FlameTail

Platinum Member
Dec 15, 2021
2,329
1,245
106
Swift
Cyclone
Typhoon
Twister
Hurricane, Zephyr
Monsoon, Mistral
Vortex, Tempest
Lightning, Thunder
Firestorm, Icestorm
Avalanche, Blizzard
Everest, Sawtooth

The legendary journey continues.
 

Doug S

Platinum Member
Feb 8, 2020
2,263
3,513
136
If they use Everest and Sawtooth they've changed from storms to mountains (or mountain ranges in the case of sawtooth)

There are plenty of storms left they could have used, maybe someone else is getting to decide the names now?
 

GC2:CS

Junior Member
Jul 6, 2018
24
19
81
If they use Everest and Sawtooth they've changed from storms to mountains (or mountain ranges in the case of sawtooth)

There are plenty of storms left they could have used, maybe someone else is getting to decide the names now?
Only thing that i saw is that A17 cores are "Everest Coll" and "Sawtooth Coll" eg. they have the same codedames as A16 cores. I could see the relation between weather -> avalanche -> mountains but honestly at some point any silicon naming scheme will break free of any meaningfull relationship.

Save for the A1 to A3 (what are those anyway) the A4-A14 feels very satysfing as I ususally get lost in some tables when comparing other chips more than a generation appart. Looks like the time has come and it is going to get complicated.

My bet is that we will see A18 with a single P core, 5 core GPU and slower neural engine (80 mm2) and then the A18 Pro with full 6 cores, 6 (7 ?) GPU cores and some really big neural engine and powerfull ISP (110 mm2).

Does that mean A18 Max = M4 ? What about A18 Ultra ?
 

FlameTail

Platinum Member
Dec 15, 2021
2,329
1,245
106
Swift
Cyclone
Typhoon
Twister
Hurricane, Zephyr
Monsoon, Mistral
Vortex, Tempest
Lightning, Thunder
Firestorm, Icestorm
Avalanche, Blizzard
Everest, Sawtooth

The legendary journey continues.

Meanwhile another legendary journey has begun elsewhere...

It started with Phoenix, and will be followed by Pegasus...
 

Tigerick

Senior member
Apr 1, 2022
656
537
106
Indeed. Also I’m curious if the 6E on the M3 Pro means anything going forward for chips that aren’t the Mx Pro. Really interested in what A18 and M4 lines are in general.
Don't expect any changes on M4 cause Apple most likely will use the same memory config, ie. 128-bit LPDDR5. Only when M5 arrive, Apple could make changes on CPU core counts.
 

FlameTail

Platinum Member
Dec 15, 2021
2,329
1,245
106
Don't expect any changes on M4 cause Apple most likely will use the same memory config, ie. 128-bit LPDDR5. Only when M5 arrive, Apple could make changes on CPU core counts.
I very skeptical that M4 will remain on LPDDR5-6400. Would be a bad move if it did.

They should upgrade to LPDDR5X-8533.

By that time even LPDDR5X/5T-9600 would be available.
 

Tigerick

Senior member
Apr 1, 2022
656
537
106
I very skeptical that M4 will remain on LPDDR5-6400. Would be a bad move if it did.

They should upgrade to LPDDR5X-8533.

By that time even LPDDR5X/5T-9600 would be available.
Do you think Apple will use LPDDR5X for one generation then change to LPDDR6 in another year?

And remember Apple has downgraded M3 Pro's memory bandwidth from 256-bit to 192-bit, why did Apple do that if they want to change?

I actually figured out on my LPDDR6 memory page, go have a look then you will understand the logic behind.

PS: That's mean Apple will bundle 8GB RAM on M4 for one more generation... :p
 
Last edited:

FlameTail

Platinum Member
Dec 15, 2021
2,329
1,245
106
Apple is never the earliest to adopt a new LPDDR standard. That has been their tradition for A series and M series SoCs.

So I expect M4 and M5 will have LPDDR5X-8533, and M6 will upgrade to LPDDR6.
 

Tigerick

Senior member
Apr 1, 2022
656
537
106
Apple is never the earliest to adopt a new LPDDR standard. That has been their tradition for A series and M series SoCs.

So I expect M4 and M5 will have LPDDR5X-8533, and M6 will upgrade to LPDDR6.
Clearly you are not familar with M1 Pro/Max launching as I stated below:
  1. At an online event on October 18, 2021, Apple announced their first MacBook Pro14/16 with M1 Pro/Max supporting latest LPDDR5-6400.
  2. Few days later, Intel announced Alder Lake which can support both DDR4/DDR5 memory.
  3. At CES 2022, AMD announced the Ryzen 6000 mobile series supporting DDR5/LPDDR5 only.
 

okoroezenwa

Junior Member
Dec 22, 2020
20
5
51
Don't expect any changes on M4 cause Apple most likely will use the same memory config, ie. 128-bit LPDDR5.
I dunno, I see them switching to LPDDR5X this year. Increased bandwidth and lower power usage seem like things they'd be interested in. Then again, they've ignored them for a bit so they could just continue this year so you might be right ¯\_(ツ)_/¯
 

Mopetar

Diamond Member
Jan 31, 2011
7,837
5,992
136
I know, and I have always disagreed. The reason everyone here believes the 18 month thing is because of a Mark Gurman tweet a few years ago, but I'm guessing he got hold of an internal roadmap showing the dates for M2 and M3 and figured out it came out to around 18 months and assumed it would be like that forever.

That could have been Apple's plan at the time. Of course they can change it later for whatever reason too and never tell anyone either way. They gain nothing by being open with that information.

They are their own customer so there's less need for them to stick to a roadmap as long as they have a good enough idea of when something will go into volume production so that they can start ramping other components, etc. in preparation.

18 months makes a good amount of sense to be a loose target for them. Even more f they don't always hit that mark, I can't see them not having any particular schedule in mind at all or deciding that each generation would be some arbitrarily different time frame for completion. If they didn't have their own internal targets, I'd expect them to align to something like TSMC's timeframes for when new nodes would be available and work around that.
 

Doug S

Platinum Member
Feb 8, 2020
2,263
3,513
136
18 months makes a good amount of sense to be a loose target for them. Even more f they don't always hit that mark, I can't see them not having any particular schedule in mind at all or deciding that each generation would be some arbitrarily different time frame for completion. If they didn't have their own internal targets, I'd expect them to align to something like TSMC's timeframes for when new nodes would be available and work around that.

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.
 
  • Like
Reactions: Tlh97 and ashFTW

FlameTail

Platinum Member
Dec 15, 2021
2,329
1,245
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
23,587
1,000
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

dr1337

Senior member
May 25, 2020
333
565
106
There's a very good reason Apple introduces new iPhones at the same time every year, and those reasons apply to other products.
Its all about volume, they ship an order of magnitude more phones than they do macs. The diminishing returns are much higher for their investments in the PC business than they are in mobile and that gives them more flexibility with the iphone.

If apple could ship 200m macs in a cycle then we would definitely see them introducing new breakthrough models every year like with iphone.
 
  • Like
Reactions: Tlh97 and moinmoin

dr1337

Senior member
May 25, 2020
333
565
106
20 mil Macs + 60-100 mil iPads per year.
Yeah so counting all three generations its still half the volume of the iphone and apple putting old CPU class chips into ipads to clear stock is not at all the same as new sales. Its not like they ship even remotely close to 60m m2 ipad pros, and you gotta consider ipad airs are still m1.

I mean why would they invest in new chips to sell ipads when clearly ipads are a market tool for them to sell old chips? If anything thats an argument that M chips should have a >18 month cadence.
 

Doug S

Platinum Member
Feb 8, 2020
2,263
3,513
136
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.

Yes and all but the last few years of that data was due to their dependence on Intel, and was one of the big reasons they wanted to take it in house. Once they did there are some obvious growing pains ramping up their teams to be able to design all the Apple Silicon stuff in addition to the iPhone SoCs they were already doing, they were affected by covid / WFH delays, departures of team members who last fully participated in the M1 generation, and TSMC's N3 troubles.

So let's revisit this in a few years and see how they're doing. I'll bet they get this stuff on a more regular cadence now that they've finally got full control of their hardware and brought Apple Silicon's onto a schedule (judging by the release of all three members of the M3 family shortly after the iPhone SoC that used the core) So long as TSMC doesn't stumble again, the Mac is going to see much more regular scheduling rather than a hodgepodge of releases at any random time like they've been forced to do in the past.
 

Doug S

Platinum Member
Feb 8, 2020
2,263
3,513
136
Its all about volume, they ship an order of magnitude more phones than they do macs. The diminishing returns are much higher for their investments in the PC business than they are in mobile and that gives them more flexibility with the iphone.

If apple could ship 200m macs in a cycle then we would definitely see them introducing new breakthrough models every year like with iphone.

Breakthrough models every year? What alternate universe are you living in where the iPhone (or any other smartphone) is providing that??

Phones are like cars, pickups and SUVs, there is a new ground up design (new body style, and internals reworked) every few years and every year there's a small tweak here and there. PCs and laptops are like heavy machinery like a Caterpillar that just have the CPU swapped out and maybe motherboard changed if required for the new CPU, but keep the identical external design with no change for years between any reworking.