Discussion Apple Silicon SoC thread

Page 73 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
23,583
996
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:

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
I don't get why people are so unhappy about external power bricks. Power supplies increase the need for cooling, that render doesn't show any fan. Is having a perfectly silent desktop worth having a power brick? Would be for me!

I wonder if the plexiglass versus aluminum top is to allow it to act as a wireless charger for something set on top of it? Though it seems like it would scratch if you set your iPhone (with its protruding lenses) on it even slightly carelessly. Maybe they'll need the "ceramic shield" for that plexiglass lol!

At least no one can complain about lack of ports here, if that render is based on reality they are even including two USB-A ports which is nice as USB-A stuff is not going away anytime soon. Heck is there a single USB-C keyboard or mouse on the market?
 

coercitiv

Diamond Member
Jan 24, 2014
6,151
11,685
136
Power supplies increase the need for cooling, that render doesn't show any fan. Is having a perfectly silent desktop worth having a power brick?
One can think of other reasons why an external power supply would make sense, but cooling ain't one. Think of it this way, if the components inside the chassis don't require active cooling, then the high efficiency PSU needed to power these components will definitely not generate enough heat to require a "system" fan. Perceived unit volume, BOM optimization, maybe faster repair times... these are all possible (though mostly debatable) reasons.

The one good reason in favor of an external power brick would be compatibility with some power standard (like USB-C) to enable a 1 cable connection to the monitor. I played around with a Chromebox using one USB-C connection to get power, data and video signal to and from the monitor, and it was great. One cable did it all. If it's great for laptops, it's great for mini computers as well.
 

Roland00Address

Platinum Member
Dec 17, 2008
2,196
260
126
I don't get why people are so unhappy about external power bricks. Power supplies increase the need for cooling, that render doesn't show any fan. Is having a perfectly silent desktop worth having a power brick? Would be for me!

The problem with power bricks is the length of the cord and how large it is. Often they can be advantageous, a power brick can be better than a generic power cable, but sometimes they are poorly designed in a way that is frustrating.

Has someone opened up these new apple 24" i-mac power bricks? Please tell me they are using GaN or a similar tech to keep the power charger smaller and generating less heat.

------

I think the main complaint people have with power bricks is when used for secondary devices and storage. Like you have to store multiple power bricks where if the brick was built into the unit it is just a universal power cord and you do not get that mental tax of making sure you do not lose a thing. Likewise with computer monitors (not computer pcs) I like the brick being built in the unit for I do not care if the monitor is half an inch thick or 3 inches thick for I can't see what is behind it, and thus the power brick is something that sits on a table, which then still has a standard size power cord that connects to the wall.
 

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
One can think of other reasons why an external power supply would make sense, but cooling ain't one. Think of it this way, if the components inside the chassis don't require active cooling, then the high efficiency PSU needed to power these components will definitely not generate enough heat to require a "system" fan. Perceived unit volume, BOM optimization, maybe faster repair times... these are all possible (though mostly debatable) reasons.

Switching power supplies have historically been inefficient enough that the industry had to do stuff like "80 plus" certification to get them make power supplies that are even 80% efficient - and often that's only at the high range of their output so if you are running it at 10% of its max output while the system is idling it might be only 50% efficient.
 

Heartbreaker

Diamond Member
Apr 3, 2006
4,222
5,224
136
Switching power supplies have historically been inefficient enough that the industry had to do stuff like "80 plus" certification to get them make power supplies that are even 80% efficient - and often that's only at the high range of their output so if you are running it at 10% of its max output while the system is idling it might be only 50% efficient.

Why expect Apple would have low quality PSU?

Most of the time they don't graph PSU efficiency down to 10%, but here is one graphed down to 10%. It's still above 70% efficient.

For an ARM Mac Mini this might be a 100 Watt PSU. So 10% = 10 Watts. 70% efficient means 3 watts of waste heat, which is irrelevant, and definitely NOT going to require active cooling.

eff.png
 

Eug

Lifer
Mar 11, 2000
23,583
996
126
If the Mac mini is indeed redesigned with an external power brick, I wonder if they'd put the Ethernet connector on the power brick, or if it will still be on the actual Mac mini. I'm thinking the latter, partially because you will be able to spec the Mac mini with 10 GigE. Or would that matter? Cuz I'm thinking the Ethernet port on the power brick is essentially likely just an extension cord adapter for hardware built into the Mac.

I don't see the Mac mini going fanless though, even with an external power supply.
 

coercitiv

Diamond Member
Jan 24, 2014
6,151
11,685
136
so if you are running it at 10% of its max output while the system is idling it might be only 50% efficient.
  • When the system is idle power consumption can drop as low as a few watts. 50% efficiency under 6W load means 3W of heat.
  • When the system is under load efficiency goes up. 85-90% efficiency under a 40W load means 4-6W of heat.
Modern external PSUs use plastic shells without ventilation holes, which should be telling for their actual cooling requirements in the first place.
 
  • Like
Reactions: lobz

Eug

Lifer
Mar 11, 2000
23,583
996
126
Ars: Covert channel in Apple’s M1 is mostly harmless, but it sure is interesting

"Technically, it's a vulnerability, but there's not much an attacker can do with it."

---

Apple's new M1 CPU has a flaw that creates a covert channel that two or more malicious apps—already installed—can use to transmit information to each other, a developer has found.

The surreptitious communication can occur without using computer memory, sockets, files, or any other operating system feature, developer Hector Martin said. The channel can bridge processes running as different users and under different privilege levels. These characteristics allow for the apps to exchange data in a way that can't be detected—or at least without specialized equipment.


 
Last edited:

B-Riz

Golden Member
Feb 15, 2011
1,482
612
136
The Apple //e was after the Apple ][ line was already firmly established. A ton of people of people had already bought the Apple ][ and Apple ][ plus. The main difference with the //e besides lower cost was that it added lower case support, but you could mod the ][ and ][ plus to support this.

I am waiting on the Mac mini. I can't get myself to buy a headless machine which can only support dual screen only if one of those can only be HDMI. I also won't buy an 8 GB Mac in 2021, because I tend to keep my Macs a very long time. In my experience, 8 GB is the sweet spot in 2021 for light usage, but memory requirements go up roughly around 50% every 5 years or so.

Ack, you got me, I was meant to reference the original Apple II. It just feels odd to get something affordable, capable, and fast that is not Intel or AMD inside and has a large company supporting the OS behind it.

I understand waiting for a more feature rich Mx product; I would consider an Mx that I could upgrade RAM and storage on if they made it available at the right price.

I am an odd duck on Apple stuff; I always get the lowest end model 6 months or more after release, either used, open box, or refurbished, only new if it is a good sale (new iPads from Micro Center are always marked down a few dollars from MSRP). The Costco deal is hard to beat as it comes with a 2 year extra warranty from Costco.

I am on the fence about needing more than 8GB for basic usage, the used 2nd gen MacBook Air has 8GB in it and has been doing great, but then I also have the beast PC for games and everything else.

If we throw the 8GB Raspberry Pi 4 in the mix, that is very capable and very affordable, but, it all comes down to knowing the limits of hardware, getting something that matches ones needs, and being an informed consumer.
 

lobz

Platinum Member
Feb 10, 2017
2,057
2,856
136
I don't get why people are so unhappy about external power bricks. Power supplies increase the need for cooling, that render doesn't show any fan. Is having a perfectly silent desktop worth having a power brick? Would be for me!
You may be right in a practical sense, but I would have expected a much more Feng shui oriented design from a lifestyle company.
 
  • Haha
Reactions: moinmoin

Heartbreaker

Diamond Member
Apr 3, 2006
4,222
5,224
136
You may be right in a practical sense, but I would have expected a much more Feng shui oriented design from a lifestyle company.

He's NOT right in the practical sense. PSU heat generation, is only a small fraction (~10%) of the heat generation of the active computer components, so it won't change a computer design from passive to active.

The reason the new iMac 24" moved to a Brick is that Jony Ive, "thin for the sake of thin" design that made the computer too thin for an integrated PSU, and even with a power brick, it still uses a fan.
 
  • Like
Reactions: Tlh97 and Gideon

lobz

Platinum Member
Feb 10, 2017
2,057
2,856
136
He's NOT right in the practical sense. PSU heat generation, is only a small fraction (~10%) of the heat generation of the active computer components, so it won't change a computer design from passive to active.

The reason the new iMac 24" moved to a Brick is that Jony Ive, "thin for the sake of thin" design that made the computer too thin for an integrated PSU, and even with a power brick, it still uses a fan.
boy you really missed the point of my post, didn't you
 

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
AMD's announcement of "V-Cache" points to another way Apple could theoretically handle GPU memory on the M1 successor. Apple has always been on the forefront of TSMC's vertical packaging technologies, so it is easy to imagine they might use it for something similar.

AMD was able to fit 32MB in 36 mm^2 in 7nm, so if Apple made a similar SRAM in 5nm at the same size I've estimated for the M1 successor of 225 mm^2 it would hold 256 MB. If it had "only" the same 2 TBps in bandwidth AMD reached, that's 2 TBps per chiplet, which would be an amazing 8 TBps in a 4 chiplet Mac Pro! Of course, for the purposes of graphics, using SRAM would be wasteful, since the reduced latency isn't necessary. If they used eDRAM (i.e. DRAM which can be made in TSMC's standard logic processes) which is around 4x as dense, they'd get 1 GB per die or 4 GB in the high end Mac Pro. I don't know the bandwidth difference between SRAM and eDRAM, but 8 TBps is kind of ridiculous and probably couldn't be fully exploited anyway.

One flaw in this plan is that 4 GB is not sufficient - Nvidia and AMD's highest end workstation GPUs have 32 or 48 GB of memory. However, Apple could stack multiple eDRAM dies per chiplet to reach whatever capacity is desired. I've figured previously based on TSMC's wafer pricing that Apple would be paying a little under $100 for a 225 mm^2 chip in 5nm, which if you add a bit to account for defects that can't be handled by redundancy we'll call exactly $100. So each eDRAM die of matching size would have similar cost - actually slightly better since the massive redundancy there should allow for essentially perfect yield but we'll still call it $100 since the math is easy. The question is, do they really need 48 GB like Nvidia's highest end workstation GPU? At essentially $100 per gigabyte that would be $4800 just in eDRAM dies!

That's a LOT more expensive than GDDR6 which as far as I can tell seems to be closer to $10 per gigabyte, so while it doesn't make much difference in lower capacities the price gap gets really large once they get enough to compare to AMD & Nvidia's top end workstation GPUs. Here's where it gets interesting though - by vertically stacking Apple can get much higher bandwidth than they could with GDDR6 which would use standard DRAM channels for communication. That might not be enough to justify the added cost, but rolling your own DRAM also allows capabilities impossible for even AMD or Nvidia to exploit - they could embed computational capability in the eDRAM array. That's been talked about for many years as a potential way to greatly accelerate certain tasks, but hasn't really been possible since commodity DRAM so thoroughly dominates the market. Perhaps in the limited case of a tightly coupled GPU and VRAM its day might finally come.

If Apple felt the cost was worth it for whatever advantage it provides, they could select a few eDRAM capacity points for the lowest to highest end Mac Pro and Macbook Pro, and vary the number of eDRAM chips stacked per chiplet/SoC. So you calculate $100 per chiplet plus $100 per GB of eDRAM, and add $100 per chiplet/SoC/eDRAM stack for testing/packaging to get a rough estimate of Apple's costs. Despite the far higher cost for eDRAM vs GDDR6 it should still stay comfortably under what Apple pays for the low/medium/high end Intel CPU + AMD GPU in the currently Mac Pro and Macbook Pro, so the cost works.

So I can't find any way to rule this out, it fits as far as cost, Apple's technical ability and willingness to use TSMC's bleeding edge packaging, and Apple's willingness to roll their own solution and not be dependent on third party silicon if they can produce something better themselves. While going a more traditional direction with commodity DRAM of some sort is more likely, if you're willing to do your own CPU and GPU, why accept a commodity solution for DRAM if you have the ability to do much better?
 

Roland00Address

Platinum Member
Dec 17, 2008
2,196
260
126
Apple can use the AMD tech just announced for it is merely TMSC’s family of 3DFabric™ technologies per the AMD keynote I just glanced over. Now TSMC 3DFabric™ is a marketing term introduced in 2020 for a web of different techs each of them with their own pro and cons for 2.5D and 3D chip stacking.

Some of the technologies are several years old, for example the 2016 Apple A10 used TSMC’s InFO (integrated fan out) and this was one of the reasons the A10 was TSMC only while previous apple chips were sometimes dual sourced like the A9. But in 2020 TSMC created a marketing name for the family of different ways to connect chips together to better integrate things.

Here is the anandtech article from that time.
 
  • Like
Reactions: Tlh97

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
Here's a link to a Google sheet that estimates die size and pricing for various interpretations of Mark Gurman / Bloomberg's rumors regarding upcoming Apple silicon chips.

Apple silicon scaling worksheet

Those numbers make no sense. They seem to think that Jade-C chop and Jade-C die are two different chips, that's obviously not true. They also have some bizarre numbers for design cost per unit, as if the 2C and 4C versions are not simply two and four Jade-C dies packaged together - which they obviously know as they reflect that in increased packaging costs.

I'm also confused by their showing both LPDDR4x/5 and HBM2 included on Jade-C die but HBM2 only on the bigger ones. Do they know something we don't, or more likely read a bad rumor somewhere? That would imply Apple has THREE types of memory on the Mac Pro since they'd need DIMMs as well. Now that would be a complex memory system, ugh!

I'll stick with my estimate of 225 mm^2 for the "Jade C" 8+2 chip, we'll see who is closer.
 

repoman27

Senior member
Dec 17, 2018
342
488
136
Those numbers make no sense. They seem to think that Jade-C chop and Jade-C die are two different chips, that's obviously not true. They also have some bizarre numbers for design cost per unit, as if the 2C and 4C versions are not simply two and four Jade-C dies packaged together - which they obviously know as they reflect that in increased packaging costs.

I'm also confused by their showing both LPDDR4x/5 and HBM2 included on Jade-C die but HBM2 only on the bigger ones. Do they know something we don't, or more likely read a bad rumor somewhere? That would imply Apple has THREE types of memory on the Mac Pro since they'd need DIMMs as well. Now that would be a complex memory system, ugh!

I'll stick with my estimate of 225 mm^2 for the "Jade C" 8+2 chip, we'll see who is closer.
There are two sets of numbers for each rumor. The first rumor is interpreted as two monolithic SoCs and then with the Pro chips as CPU + discrete GPU. The second rumor is interpreted first as four monolithic SoCs and then as a universal tile followed by the 4 variants as packages built using that one tile.

Jade C-Chop won't have HBM, just LPDDR4X/5. Jade C-Die with 32 GPU cores requires either GDDR or HBM, and according to the Bloomberg rumor, will be available with up to 64GB of memory. There is no way that four 16 GB HBM2E stacks makes sense with a single tile. The memory architecture will continue to be unified, however it will be tiered: two LPDDR4X/5 packages plus two HBM2E stacks. The two and four tile packages will be able to take 4 or 8 HBM2E stacks and will forgo LPDDR entirely. Instead, they will support DDR5 RDIMMs with ECC to enable scaling up to 1 or 2 TB of DRAM.

Apple needs to support all three memory technologies to make this work. Once they add a controller to a tile, it's along for the ride whether it is exposed by the final package or not. Each tile will therefore have two LPDDR4X/5 controllers that each support 4 x16 channels, 2 HBM2E controllers with 1024-bit interfaces, and a DDR5 controller that supports 4 x32 channels.

edit: This actually represents a *simplification* of the memory system compared to the Intel Macs these are replacing, which currently support between 3 and 6 independent memory subsystems: LPDDR or DDR for the CPU and IGP (if present), separate LPDDR for the T2 or storage controller, plus one GDDR or HBM system for each discrete GPU.
 
Last edited:

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
They probably want a separate event for that. I've also said for months that they would want to use the A15 cores and N5P for the higher performance Macs, and since N5P mass production only ramped in April they might need a few more weeks to be ready to ship.

Their real "deadline" is having them ready for back to school, which means an announcement by end of July.
 

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
M1's are plenty powerful for back to school.

The real deadline is "when they are ready".


There is only one ARM MBP, the low end model at the smallest size. You don't think they want ARM Macs across the whole laptop line for back to school?

Now Mac Pro on the other hand can wait, no one is getting those for school.