Discussion Apple Silicon SoC thread

[Eug]

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:

Page 78 - Discussion - Apple Silicon SoC thread

Page 78 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M1 Ultra discussion here:

Page 109 - Discussion - Apple Silicon SoC thread

Page 109 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M2 discussion here:

Page 127 - Discussion - Apple Silicon SoC thread

Page 127 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

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:

Page 215 - Discussion - Apple Silicon SoC thread

Page 215 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M4 Family discussion here:

Page 263 - Discussion - Apple Silicon SoC thread

Page 263 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

 

[trivik12]

https://twitter.com/x/status/1719150026419490965

Its interesting the M3 pro and base version of M3 Max regressed in core counts. Only Apple can call it a Yuge upgrade Cant wait for independent benchmarks comparing it to M2 series.

 

[repoman27]

https://twitter.com/x/status/1719150026419490965

Its interesting the M3 pro and base version of M3 Max regressed in core counts. Only Apple can call it a Yuge upgrade Cant wait for independent benchmarks comparing it to M2 series.

Well, I'm not sure how many people who bought a 14 or 16-inch MacBook Pro with an M2 Pro or Max since they were released this past January are really looking to upgrade already.

 

[trivik12]

Well, I'm not sure how many people who bought a 14 or 16-inch MacBook Pro with an M2 Pro or Max since they were released this past January are really looking to upgrade already.

Its not about them upgrading. Why should anyone buy M3 based macs when they can buy M2 based macs with more P cores and higher bandwidth memory and also more GPU cores in some cases. Plus M2 Pro will probably be on sale making it even cheaper.

 

[Eug]

Its not about them upgrading. Why should anyone buy M3 based macs when they can buy M2 based macs with more P cores and higher bandwidth memory and also more GPU cores in some cases. Plus M2 Pro will probably be on sale making it even cheaper.

Longevity for OS updates. The M3 models will get more updates going forward. Plus there are some codec upgrades.

 

[adroc_thurston]

Apple did three separate designs, rather than two plus a chop.

They've been doing three separate designs since M1.

Where exactly are you seeing cost cutting here?

M3 Pro got gutted.

And they're shipping a ~400mm die on TSMC N3 less than 12 months after reaching volume production.

TSM has been good at ramping nodes into volume since N7.
Not really an achievement?

 

[trivik12]

Longevity for OS updates. The M3 models will get more updates going forward. Plus there are some codec upgrades.

you think getting another year of upgrade support is worth giving up as much as 3P for M3 Pro base model !!! Apple supports 8 to 10 years of OS support for its MBP and so I am not sure the additional year is worth anything is people upgrade their laptop well before that period. And Code Upgrades !!!!

 

[FlameTail]

Geez. M3 Pro got shafted.

Only 6P cores and 192-bit memory bus?

 

[FlameTail]

I was expecting two things, both of which were not met:

1. LPDDR5X RAM

2. 20% more GPU cores - Presuming M3 family is based on the A17 foundation. A17 got 20% more GPU cores (5 -> 6).

 

[adroc_thurston]

Only 6P cores and 192-bit memory bus?

And one less GPU core.

 

[FlameTail]

Also how much L2 / System cache do these have?

 

[FlameTail]

Apple M2 Die Shot and Architecture Analysis – Big Cost Increase And A15 Based IP

Apple announced their new 20 billion transistor M2 SoC at WWDC. Unfortunately, it’s quite a minor uplift in performance in some areas such as CPU. Apple’s gains mostly came from the GPU and video editing side of things.

www.semianalysis.com

Apple's M1 Pro, M1 Max SoCs Investigated: New Performance and Efficiency Heights

www.anandtech.com

According to Semianalysis M2 had 8 MB SLC. But according to Anandtech it's 16 MB SLC.

Who is right?

 

[Doug S]

I'm still surprised they released the base, Pro and Max all at once, and are shipping at the same time (Max a few weeks later but basically the same time) I wonder what the reason was for Pro not being a chop of the Max this time around? Maybe they couldn't make the chop work while arranging it so the Max could get three way connectivity to do a 4 SoC "Extreme" version?

The memory configurations are....weird. Maybe they are binning not only on working CPU and GPU cores, but also on working memory controllers? With the reported yields for N3B they may have had to make some compromises to get enough working 92 billion transistor chips out the door. We'll have to see if they do the same thing for M4 which would presumably be made on N3E and not have to deal with yield issues.

 

[FlameTail]

Yeah why do they only bin CPU and GPU.

Don't the NPU, caches and RAM controllers get defects?

 

[GC2:CS]

This was an interesting launch.

The A17 Pro and M3 diverge on features. (?) A17 has a beefier neural engine 2x as big, while M3 got the dynamic caching on the GPU which from what I understand A17 Pro does not have or does not have enabled.

Then we have M3 Pro which is 1,5x M3 not 2x M3. It seems there is effort to make it cheaper entry level for the Pro than the previous gen.

Then we have binned memory buses in addition to CPU and GPU. Aggain M3 Max seems huge with 5/6 the transistors of an M1 Ultra. No wonder the bins have to be more aggresive on a new process

We have very little comparison to M2 generation, focusing instead on comparing to M1.

From the graphs we saw, the power consumtion is slighthly higher M1 vs M3 on CPU but that was the case for M2 as well. We have no info on efficiency improvements compared to M2. We have no battery life improvement compared to M2 gen.

Would like to see the GPU curves compared, how much more power the max Max die consumes - a guess is 5x12 + 4x0,5 = 65-70 W for the CPU and 50 W for the GPU. How well can the 14" cope with a full blown Max at max load ?

Is there anything useful on those die shots to be seen ? The memory bus (?) on the sides of the M3 Max seems particullary blown up compared to other chips.

Is there any way to compare the costs of configurations gen on gen ? I got completely lost and confused on configurations site this time.

Is the M4 generation coming like this next year ?

 

[Henry swagger]

https://twitter.com/x/status/1719090003261677629

M3 better perform well in sales because the m2 floped hard.. apple and Qualcomm have very small market share compared to other oem's

 

[GC2:CS]

https://twitter.com/x/status/1719090003261677629

M3 better perform well in sales because the m2 floped hard.. apple and Qualcomm have very small market share compared to other oem's

Well it simply looks like Apple silicon Macs put people rushing to upgrade but not brought many new consumers from the windows side.

Now we are left with tens of millions of M1 users who are excessively happy with their cheap purchase and will remain so for years to come. While windows users do not see Apple silicon as a big enough incentive to switch.

This might turn into Apple lowering their standards on the M processors as we saw yesterday. Designing an A series means an infinitely more return in investment compared to designing 3 very expansive chips with far less sales.

 

[Henry swagger]

Well it simly looks like Apple silicon Macs put people rushing to upgrade but not brought many new consumers from the windows side.

Now we are left with tens of millions of M1 users who are excessively happy with their cheap purchase and will remain so for years to come. While windows users do not see Apple silicon as a big enough incentive to switch.

This might turn into Apple lowering their standards on the M processors as we saw yesterday. Designing an A series means an infinitely more return in investment compared to designing 3 very expansive chips with far less sales.

yeah well said

 

[FlameTail]

Perhaps Apple should release a new generation Apple M chips once every 2 years. That would be better for their economics and align with the A series chips, unlike the current bizarre 18-month cadence.

 

[jpiniero]

TSM has been good at ramping nodes into volume since N7.
Not really an achievement?

This is presumably N3B though. Who knows how long it will be around. Maybe it will be around longer than I think.

 

[FlameTail]

https://x.com/t3mporarybl1p/status/1719317658255110416?s=20

Bizarre NPU stuff

 

[jpiniero]

Bizarre NPU stuff

Cost cutting by saving space.

 

[Heartbreaker]

This might turn into Apple lowering their standards on the M processors as we saw yesterday. Designing an A series means an infinitely more return in investment compared to designing 3 very expansive chips with far less sales.

This was why I always had a hard time seeing Apple doing the ARM transition for Macs. Macs have MUCH less volume than iPhones, and they needed more chips to cover the range of Macs. It seemed like chip economics were really against them to doing ARM Macs. But they did do it...

 

[igor_kavinski]

But not surprising IMO.

Guess most people are using their MBP for basic tasks, for a lot more money than something like a $500 laptop with 8GB/256GB. Apple is lucky to have such users. I'm guessing at best, MacOS leaves 6 GB RAM available for application usage?

 

[Heartbreaker]

Then we have M3 Pro which is 1,5x M3 not 2x M3. It seems there is effort to make it cheaper entry level for the Pro than the previous gen.

Also makes the Max chip a more significant upgrade over the Pro. Previously the Pro and Max chips shared the same CPU core count, so for people unconcerned with GPU performance, there was really no point going for the Max chips, so I bet Max chips were not big sellers.

Now Max chips have double the CPU Performance cores of Pro chips, so easier to upsell Max chips.

We have very little comparison to M2 generation, focusing instead on comparing to M1.

We largely have diminishing returns across the chip industry, such that previous generation comparisons look lackluster, for the majority of these tech products. So we often see "skip a generation" comparisons to show significant progress.

Though from my perspective it seems like they would be better served spacing out the generations a bit more, to get more significant intergenerational gains, and save redundant chip design costs.

Leave the rapid incremental upgrades for the high volume smart phone markets.

 

[Tup3x]

https://twitter.com/x/status/1719090003261677629

M3 better perform well in sales because the m2 floped hard.. apple and Qualcomm have very small market share compared to other oem's

I don't see M3 doing any better.