Discussion Apple Silicon SoC thread

Page 260 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
23,825
1,396
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:

SpudLobby

Senior member
May 18, 2022
989
680
106
“Well the only time they delay more than precisely two weeks is if it’s a new product announcement or it’s part of another broader announcement, or other products are released”

Like yeah it’s not a terrible rule but it also quickly becomes laughable when the margins we’re talking about here are… another two weeks on top of that.

OLED on the iPad with a new haptic feedback Magic Pen or whatever and an M4 also potentially makes it less ridiculous
 

SpudLobby

Senior member
May 18, 2022
989
680
106
Actually, we’re all offtrack here. If Gurman speculates a larger Neural Engine from M2 -> M4 is half the show due to new AI features it can take (further if not unique qualitative stuff) advantage of, the big problem with this is Apple’s iOS 18 is what is supposed to deliver exactly that on the software side.

WWDC is in June.

And before someone says “APPLE DOES USE AI DUDE” I am aware of that but it simply isn’t as extensive as what’s coming in iOS 18 at all with assistants for notes/reminders (could be big for art with the pencil or note taking too), Safari assistant, a Siri upgrade, etc.



So that is strange.
 
  • Like
Reactions: carancho

repoman27

Senior member
Dec 17, 2018
381
536
136
Actually, we’re all offtrack here. If Gurman speculates a larger Neural Engine from M2 -> M4 is half the show due to new AI features it can take (further if not unique qualitative stuff) advantage of, the big problem with this is Apple’s iOS 18 is what is supposed to deliver exactly that on the software side.

WWDC is in June.

And before someone says “APPLE DOES USE AI DUDE” I am aware of that but it simply isn’t as extensive as what’s coming in iOS 18 at all with assistants for notes/reminders (could be big for art with the pencil or note taking too), Safari assistant, a Siri upgrade, etc.



So that is strange.
Maybe not.

Apple does these types of spring launches because the announcement might not fit into the WWDC keynote, or they want to get the products out ahead of the conference. I think this is true on both counts with these new iPads. Too much to talk about, and a lot of it isn't focused on things developers care about. The "Let Loose" event will be about the hardware, and software will take a back seat. If the M2 is going to be the torch bearer for the new neural engine and AI features though, Apple wants it to be available before WWDC, so developers can have one in hand when the first iOS / iPadOS 18 betas drop. The WWDC keynote will focus on the software, all the new APIs and OS features that can leverage the hardware that has already been announced or is clearly in the pipeline.
 
  • Like
Reactions: carancho

Doug S

Platinum Member
Feb 8, 2020
2,784
4,744
136
The M2 Air was unveiled at WWDC and launched a month later. People exaggerate how close to release Apple is sometimes, it varies. The key with them is when they do officially release, it isn’t a paper launch. Whereas with other vendors you’ll see “releases” sometimes and it’s a paper launch.

iPhones also have had delays before for the Pro or Max models.


“In fact, Apple isn't launching all four iPhone 12 models at the same time. Only the iPhone 12 and iPhone 12 Pro will be available to preorder on Friday, Oct. 16, with deliveries a week later on Oct. 23.The iPhone 12 Mini and iPhone 12 Pro Max will go on sale Nov. 6, and start shipping on Nov. 13.“

This isn’t rare for Apple. They might do 2-6 week delays for a real launch from announcement but when it releases, it *hits in volume*.

Apple also isn't shy about having new products for which they can't fully satisfy demand, or at least they weren't in the past. They've got better as the production scale as increased over the years, but the shelves in Apple stores used to be empty after a new iPhone launch. If you didn't get your order in within a few hours of when they started taking orders you might see instead of next week your ship date is a month or two out. Even now you might see particular SKUs like "purple iPhone 14 Pro Max 256 GB" be a month out while other SKUs are readily available, because they can only guess exactly the relative product mix customers will demand and when they're wrong temporary shortages result until they tweak production targets to satisfy the demand for things they underestimated.

They may choose when to launch based on "we want to be able to ship 1 million on day one and 2 million in the first month" (or whatever, I have no idea what volume iPad Pro sells) and when they can forecast that they pull the trigger on the announcement and announce a ship date that will make that goal come true.

They may not even be gated by Apple Silicon, and none of this has to do with TSMC at all. They are using those fancy dual stack OLEDs no one else is so it is quite possible that will be the biggest constraint on how many they can ship over the next few months.
 

SpudLobby

Senior member
May 18, 2022
989
680
106
Apple also isn't shy about having new products for which they can't fully satisfy demand, or at least they weren't in the past. They've got better as the production scale as increased over the years, but the shelves in Apple stores used to be empty after a new iPhone launch. If you didn't get your order in within a few hours of when they started taking orders you might see instead of next week your ship date is a month or two out. Even now you might see particular SKUs like "purple iPhone 14 Pro Max 256 GB" be a month out while other SKUs are readily available, because they can only guess exactly the relative product mix customers will demand and when they're wrong temporary shortages result until they tweak production targets to satisfy the demand for things they underestimated.

They may choose when to launch based on "we want to be able to ship 1 million on day one and 2 million in the first month" (or whatever, I have no idea what volume iPad Pro sells) and when they can forecast that they pull the trigger on the announcement and announce a ship date that will make that goal come true.

They may not even be gated by Apple Silicon, and none of this has to do with TSMC at all. They are using those fancy dual stack OLEDs no one else is so it is quite possible that will be the biggest constraint on how many they can ship over the next few months.
I agree with that, Doug, though it’s possible the demand gate is incorporated into the delay, e.g. entirely possible the 1 million threshold is June. They’re not shy about that either.

This whole thing is kinda ridiculous. They absolutely could do one month (no more though). They may not.

More interesting is: is it the M4.
 

poke01

Platinum Member
Mar 8, 2022
2,205
2,802
106
I agree with that, Doug, though it’s possible the demand gate is incorporated into the delay, e.g. entirely possible the 1 million threshold is June. They’re not shy about that either.

This whole thing is kinda ridiculous. They absolutely could do one month (no more though). They may not.

More interesting is: is it the M4.
The iPad Pro is a popular device. Many people and businesses buy them. It could easily sell a million in a month depending on stock.

As for the chip at this point going by the identifiers it’s anything but the current M3 chip. Apple refreshes the iPad Pro every 18 months. The last refresh was November 2022 and so it’s been around 18 months.

There’s no point beating the bush, only 6 more days and if the M4 is released next week, it will be a fun week.
 

SpudLobby

Senior member
May 18, 2022
989
680
106
The iPad Pro is a popular device. Many people and businesses buy them. It could easily sell a million in a month depending on stock.

Sure.
As for the chip at this point going by the identifiers it’s anything but the current M3 chip. Apple refreshes the iPad Pro every 18 months. The last refresh was November 2022 and so it’s been around 18 months.

There’s no point beating the bush, only 6 more days and if the M4 is released next week, it will be a fun week.
 

repoman27

Senior member
Dec 17, 2018
381
536
136
We're definitely getting into the weeds here, but several of these numbers were brought up in relationship to whether or not a new SoC introduced at the event on Monday could be manufactured on N3E or not.

While there is no conclusive answer to that question, there are some very clear upper and lower bounds regarding the timing. We know TSMC N3E entered volume production some time between the first week of October and the last week of December 2023. We also know Apple's product announcement is scheduled for May 7. Given that WWDC is scheduled to start on June 10, the most likely window for new product availability is between May 7 and May 31. It is incredibly unlikely that Apple would plan an announcement for May 7 and release products earlier than that or after the final Friday in May given the timing of WWDC. That leaves a production window for a device with an N3E based SoC of not more than 8 months (totally doable) but possibly as short as 5 months (probably not doable). We're gonna have to wait and see, and Apple may not make any mention of the manufacturing process they're using other than referring to it as "3-nanometer".

As for iPad sales figures, unit sales for calendar Q2 have been hovering around 12 million, ticking up to around 14 million in Q3, and of course new product introductions like this tend to give those numbers a boost. The vast majority of iPads sold in any given quarter are going to be the less expensive models based on the A-series SoCs. I would expect the new iPad Pros to account for roughly 20% of the mix during the period directly following the launch though, hence my estimate of a 14-week inventory target of 2.5 million units (20% of 12.5 million).
 

mikegg

Golden Member
Jan 30, 2010
1,844
467
136
The phrase "whatever you're smoking" is a colloquialism to indicate that a person finds a claim so wild or absurd as to believe you must be under the influence of drugs to make it. It's hardly unfriendly in and of itself and they're are considerably more rude ways of conveying that sentiment.

But the post you linked is over three and a half years old. What kind of grudge are you carrying that you would even dredge something like that up to try to prove a point in the first place?

Also, in reading that post and what was being replied to, it doesn't really make you look any better. We're about 18 months away from being able to actually evaluate your hypothetical (or whatever you would care to call it) of Apple hitting 50% PC market share. I don't think that's going to happen. They've only been above 20% in the US in a single quarter based on data from a quick Google search.

So I think it's quite fair for the other poster to have used that particular phrasing. Your prediction was wildly speculative and if you had bet money on it you would have lost. Even if that weren't the case, your own post was no less inflammatory in terms of language used, so I would not consider you to be some innocent victim receiving misdirected derision. I don't care to read back through three year old posts to see who started it, but it's my opinion that any curt replies you received were deserved.
You're sounding more and more ridiculous.
 

mikegg

Golden Member
Jan 30, 2010
1,844
467
136
Perhaps we get something like an M3X where it's not an M3 but is not an M4 either.
I don't understand why people here have such a hard time believing that Apple could release a new core in an M chip rather than an A chip in special circumstances.

A18 Pro is already in production at TSMC because Apple always starts production of iPhone chips in Spring. Therefore, the design work for A18 Pro has long been finished which means given enough early planning, Apple could have gotten the M4 ready by May. It's not farfetched given that they planned these things years in advance and N3B was known to be a transitional node.

Combine the fact that Apple is behind in GenAI applications and stock investors have not rewarded AAPL because they're so behind, it's not surprising to me that Apple is being aggressive with the M4 generation.
 

SpudLobby

Senior member
May 18, 2022
989
680
106
I don't understand why people here have such a hard time believing that Apple could release a new core in an M chip rather than an A chip in special circumstances.
Oh, I totally believe this is possible. They’re not wedded to a core order of release.

My more interesting question is: does Apple even have juice left in the tank on perf/GHz? Pace has been slow albeit they still lead by 10-15% over like the X4 or Nuvia.

But the actual performance/power curve for ST with Apple is even further ahead and is what is still so impressive, even if they’re pushing it too high, it’s still a different league.

The 8 Gen 3’s X4 matches the A14 in SpecInt and SpecFP (or + 5%) at nearly 40% more power. That’s N5 vs N4P too, and LPDDR4x vs LPDDR5x wrt power and perf impacts.

On GB5, it’s [8 Gen 3] more like +7% and basically M1-caliber (not sure on power though but likely similar vs A14, 4.1 vs 5.7 watts was the difference).

Obviously like AMD and Intel can’t even get that kind of single thread power compression for now, and I’d be surprised if Lunar Lake substantially beat it (the curve, not absolute perf) — and if it does just match that kind of sub-10W curve from QC or Apple without blowing it out, doing so on N3B with 140mm^2 is just sad lol.
 

poke01

Platinum Member
Mar 8, 2022
2,205
2,802
106
Obviously like AMD and Intel can’t even get that kind of single thread power compression for now, and I’d be surprised if Lunar Lake substantially beat it (the curve, not absolute perf) — and if it does just match that kind of sub-10W curve from QC or Apple without blowing it out, doing so on N3B with 140mm^2 is just sad lol.
What’s the size of the base M3 die?
 

Eug

Lifer
Mar 11, 2000
23,825
1,396
126
What’s the size of the base M3 die?
I summarized the die size estimates for the various M series chips here:


M3 is supposed to be around 146-ish mm2 on N3B. >140 anyway. That could suggest a hypothetical M3 on N3E might be over 150 mm2.
 
  • Like
Reactions: poke01

repoman27

Senior member
Dec 17, 2018
381
536
136
What’s the size of the base M3 die?

I summarized the die size estimates for the various M series chips here:


M3 is supposed to be around 146-ish mm2 on N3B. >140 anyway. That could suggest a hypothetical M3 on N3E might be over 150 mm2.
My initial estimates based on the die images provided by Apple and measurements of IP blocks shared with the A17 Pro were as follows:

M3: 12.64 mm x 10.67 mm = 135 mm²
M3 Pro: 12.69 mm x 14.96 mm = 190 mm²
M3 Max: 20.40 mm x 21.03 mm = 429 mm²

I'm pretty sure these are a lot closer than the ones Eug sourced, but they are still based on A17 Pro measurements provided by Revegnus (@Tech_Reve). I haven't had the opportunity to verify them yet though.
 

Eug

Lifer
Mar 11, 2000
23,825
1,396
126
My initial estimates based on the die images provided by Apple and measurements of IP blocks shared with the A17 Pro were as follows:

M3: 12.64 mm x 10.67 mm = 135 mm²
M3 Pro: 12.69 mm x 14.96 mm = 190 mm²
M3 Max: 20.40 mm x 21.03 mm = 429 mm²

I'm pretty sure these are a lot closer than the ones Eug sourced, but they are still based on A17 Pro measurements provided by Revegnus (@Tech_Reve). I haven't had the opportunity to verify them yet though.
I didn't do any calculations myself, but FWIW, the 146 number was published by Tom's Hardware, referencing a tweet from @Frederic_Orange who counted 415 M3 dies for a 3 nm wafer:



Perhaps going about it this way overestimates the size, but I did not attempt to work out the math, since I wouldn't know how to calculate the unused area of the wafer.

Anyhow, if it truly is 135 mm2 on N3B, then that could put a hypothetical N3E version of M3 at well over 140 mm2. I personally won't attempt to guess the size of an M4 on N3E.
 

Eug

Lifer
Mar 11, 2000
23,825
1,396
126
I don't know how accurate this would be, but knowing that the width of the chip is roughly 18-20% wider than the height, I plugged some numbers into this die per wafer calculator, using the default settings for the last 3 variables:

13.22 x 11.11 mm = 146.87 mm2 = 415 dies

Screenshot 2024-05-03 at 12.02.53 AM.png
 
  • Like
Reactions: Mopetar

SpudLobby

Senior member
May 18, 2022
989
680
106
What’s the size of the base M3 die?
146mm^2 lololol.


So the same thing.

Edit: or 135. Whatever. It’s the same and it’s gonna be funny because it’ll show all the node whiners that it doesn’t do everything. Very much doubt Intel is going to pull a 3150 GB6 ST under 10W, and we already know the GPU is more M1-class on power and M2 on perf at peak (well slightly higher with more power) by their own documents.

It will be a good part because they’ll fix some idle stuff relative to current Intel parts and likewise Lion Cove and Skymont will be big upgrades and on a new node, but again, directional movement getting them to good enough I think is the best way to see it. I might even buy it.

But I don’t think it seems impressive from an engineering POV at all by any accounts, it’s more like Intel showing up to class halfway through. A very late catchup with more $ spent on it.
 
Last edited:

Tigerick

Senior member
Apr 1, 2022
696
602
106
I summarized the die size estimates for the various M series chips here:


M3 is supposed to be around 146-ish mm2 on N3B. >140 anyway. That could suggest a hypothetical M3 on N3E might be over 150 mm2.
Eug, you do know the advantage of N3B process. Have you ever thought about what Apple going to upgrade the upcoming M4 SoC?

I am expecting 2 more e-cores, 2 more GPU cores and SLC cache expansion. Not to mention the upgrades of NPU TOPS. Why does Apple switching to different design and lower density process for Mac lineup? I am sure Apple will put in more than 100 billion transistors in upcoming M4 Max. That's why I believe Apple will maintain the same process for upcoming M4 lineup...And that's what Mark Gurman claimed as well
 

Eug

Lifer
Mar 11, 2000
23,825
1,396
126
Eug, you do know the advantage of N3B process. Have you ever thought about what Apple going to upgrade the upcoming M4 SoC?

I am expecting 2 more e-cores, 2 more GPU cores and SLC cache expansion. Not to mention the upgrades of NPU TOPS. Why does Apple switching to different design and lower density process for Mac lineup? I am sure Apple will put in more than 100 billion transistors in upcoming M4 Max. That's why I believe Apple will maintain the same process for upcoming M4 lineup...And that's what Mark Gurman claimed as well
? Are you suggesting the next chip series will also be N3B? If so, you're pretty much the only person that believes this, and no, Mark Gurman didn't claim this.
 

Tigerick

Senior member
Apr 1, 2022
696
602
106
? Are you suggesting the next chip series will also be N3B? If so, you're pretty much the only person that believes this, and no, Mark Gurman didn't claim this.
You should check the leaks by Mark Gurman
The M4 chips will be built on the same 3-nanometer process as the M3 chips, but Apple supplier TSMC will likely use an improved version of the 3nm process for boosts in performance and power efficiency
 

Eug

Lifer
Mar 11, 2000
23,825
1,396
126
You should check the leaks by Mark Gurman

--

The M4 chips will be built on the same 3-nanometer process as the M3 chips, but Apple supplier TSMC will likely use an improved version of the 3nm process for boosts in performance and power efficiency
I think you are misunderstanding that sentence. N3B and N3E are both N3 "3 nm", but N3E is an improved version of the 3 nm process for boosts in performance and/or power efficiency. According to the analysts and the experts here, N3B is essentially dead soon.

In fact, that's probably partially the reason there was a rumour last year that claimed that Apple would actually move its existing N3B chips to N3E, although many here said it isn't likely because the redesign costs would be prohibitive.

I don't know how accurate this would be, but knowing that the width of the chip is roughly 18-20% wider than the height, I plugged some numbers into this die per wafer calculator, using the default settings for the last 3 variables:

13.22 x 11.11 mm = 146.87 mm2 = 415 dies

View attachment 98270
I tried running the numbers again, this time with @repoman27's numbers, and this is the result:

12.64 x 10.67 mm = 134.87 mm2 = 458 dies.

Screenshot 2024-05-03 at 9.15.20 AM.png

However, as mentioned, Apple's own image showed 415 dies.

BTW, my 13.22 x 11.11 mm guesstimate from working backwards assumed that the width is 19% longer than the height. @repoman27's 12.64 x 10.67 mm puts the width at 18.5% longer than the height, so we're in the same ballpark there.
 
Last edited:
  • Like
Reactions: Ghostsonplanets

repoman27

Senior member
Dec 17, 2018
381
536
136
The reason I believe my measurements are more accurate is because the source image that @Frederic_Orange started with is problematic. It's not really a photo of a wafer, it's a screen capture of an animation taken from the Apple keynote. That might be an undoctored yet dramatically lit video of an actual M3 wafer, or it might not. Let's say that it is. Just counting the gross die per wafer and plugging numbers into a die per wafer calculator until you get a match is not the best way to go about it—there are way too many variables involved.

However, we do know that the wafer itself is exactly 300 mm in diameter. So you can just count the maximum number of whole dies per row / column, add 1, and divide 300 by that number to arrive at a decent ballpark of the dimensions. In this case that works out to 13.6 mm x 11.5 mm = 156 mm², which isn't too far off. Measuring the dies themselves is more accurate.

I went frame by frame through the official version of Apple's "Scary Fast" event video in the highest resolution available and took a screen grab that I felt best showed the complete wafer. I opened that image in Photoshop and used the measurement tool with the measurement scale set to reflect the full diameter of the wafer being 300 mm. I then exactly measured the tallest column and widest row of complete dies, divided by the number of dies, and adjusted those numbers slightly for saw kerf. That yielded 13.3 mm x 11.1 mm = 148 mm². Despite my best efforts at avoiding error caused by foreshortening due to perspective, the ratio of those measurements does not exactly match the aspect ratio we find in Apple's M3 die shots. We also have no way of knowing if Apple stretched or altered the aspect ratio of any images provided in their marketing materials. The final issue is that measurements taken from a wafer include the full seal ring and scribe lanes, whereas Apple's die images appear to include some of the seal ring but are clearly cropped in slightly from what would be the die edge, so measurements taken from the die shots will be smaller than those based on wafer images.

The results from the two methods differ by a little less than 10%, so not a huge amount. However, if you want numbers that are comparable to the die sizes provided in the past by reverse engineering firms like TechInsights, I still believe the lower estimates I provided earlier are probably closer. I would love for TechInsights, Yole Group / SystemPlus, or Techanalye to throw us a bone and give us actual measurements for at least one of the M3 family chips.
 

mikegg

Golden Member
Jan 30, 2010
1,844
467
136
My more interesting question is: does Apple even have juice left in the tank on perf/GHz? Pace has been slow albeit they still lead by 10-15% over like the X4 or Nuvia.
No one knows besides the people on the chips but I'm guessing yes. I'm going to guess that the M4 generation will have a surprising improvement in IPC.

The CPU and GPU were both significantly redesigned but did not produce a high IPC improvement over A16. I'm guessing that Apple wanted to set the stage for optimizations after doing the hard work of redesigning the cores along with a brand new node.