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

 

[amrnuke]

I wouldn't sell a cheaper Macbook. Apple either has to sacrifice on features, margins, or both to move into a lower price bracket. A 12" screen might save $10 on the BOM. That doesn't translate into a $200 retail price drop.

Apple HAS sold a cheaper, smaller Macbook Air before, and it sold well, so it's not really relevant whether YOU would sell a cheaper Macbook - since you replied to my question, I asked "how" you would differentiate it, not "whether" you would. If you want to be contrarian rather than constructive, please keep it to yourself.

 

[Heartbreaker]

Apple HAS sold a cheaper, smaller Macbook Air before, and it sold well, so it's not really relevant whether YOU would sell a cheaper Macbook - since you replied to my question, I asked "how" you would differentiate it, not "whether" you would. If you want to be contrarian rather than constructive, please keep it to yourself.

It has to exist before you differentiate it.

Whether you like it or not I gave you a reasonable answer. I don't think a significantly cheaper Macbook is coming, because Apple won't sacrifice margins or quality to get there.

As I said before, I think $899 is a likely floor for Macbook pricing (IIRC that is the cheapest Macbook price ever). How they get to that price, simply keep selling the older model when a new one comes out. Fully amortized design and parts, no quality sacrifice, maybe a tiny hit to margin. It would be differentiated by being older, slower and cheaper.

 

[lobz]

It is Apple's answer to the low-cost market, but I don't think it's truly ready yet. Remember, an iPad Air with a Magic Keyboard is just shy of $900... and there is no Magic Keyboard for the base iPad. You don't need a keyboard, but we're talking about PC replacements, and many people aren't ready to go touch-only.

You're also overselling the iPad's capabilities a bit. It's much better than it was even a couple of years ago, but it's not at the point where it's great for serious productivity yet. I can't do my job on one, and not because of apps — our web content system won't play nicely with it. I don't think Apple needs to do "macOS with touch," but iPadOS needs to be a more generalized platform to really get people to ditch their PCs.

It's not people that aren't ready to go touch only, but the devices. I'm typing on touchscreens since the first iphone came out and no matter how much better they are now, I still hate and avoid it when possible. I even hate typing on 99.99% of all laptop keyboards for more than 10 minutes. Freaking insufferable.
I'm really happy for all the creative professionals out there, but not everyone's so lucky that they can work with more pictures than words.

 

[Commodus]

It's not people that aren't ready to go touch only, but the devices. I'm typing on touchscreens since the first iphone came out and no matter how much better they are now, I still hate and avoid it when possible. I even hate typing on 99.99% of all laptop keyboards for more than 10 minutes. Freaking insufferable.
I'm really happy for all the creative professionals out there, but not everyone's so lucky that they can work with more pictures than words.

I suspect a lot of people won't go touch-only at all, but you're right that it's difficult to go touch-only in the current environment. The iPad has the best on-screen keyboard in my experience, and even then...

 

[Mopetar]

Yeah, this part of the discussion is a lost argument that is dragging out. I can see Apple knocking $100 off the price of a laptop, but that is about it. A large part of the PC market is low end trash, that Apple want's nothing to do with.

You're exactly right. At a certain point it isn't worth it to try to go after more market share. If they made $600 on each of those $1000 MacBooks, dropping the price $100 requires them to sell at least 20% more MacBooks than the usually do just to make the same profit. That's certainly possible, but once you start pushing past that price level, the numbers start to make less and less sense, particularly considering that Apple would need to invest a lot of money into the additional infrastructure just to produce 20% more computers since they do some precision work on their bodies and likely have pretty strict tolerance levels as opposed to the companies that are just dumping components in cheap plastic that barely fits together.

They can build towards doing something like that slowly without running into too much trouble. However, once they start needing to produce something like three times as many computers they're going to need to spend a lot of money to have factories built to produce what they want. They certainly have the money to do that, but it's a lot of trouble to go through just to break even, never mind the initial investment. Granted the calculation is simplistic and doesn't consider fixed cost, but it's easier to look at it from the other direction. Getting to 50% market share would require Apple to sell around six to seven times as many computers as they do now.

 

[Roland00Address]

Side question for this thread that is actually relevant to this thread.

Remember I posted earlier how this 2020 Macbook Air / Pro / Mac Mini is the 2008 dream of the first Macbook, a dream that started being realized in 2010 with the revision 2.

Well another way to put it, it is the AMD dream of a "fusion" chip where you use the CPU, the GPU, and other things on a single piece of silicon using that part of the chip to best do the work for the task you need.

But when did AMD start using "fusion" marketing for I can't recall? The first fusion chips were the low powered 40nm intel atom competitors Brazos followed quickly later that year with Llano which was 32nm. But when did the marketing happening with AMD slides and so on*?

That is the question!


Edit: Anandtech has an article preview from 2010 with the term Fusion as one of his tags, but is there anything earlier than Nov 2010.

 

[NostaSeronx]

But when did the marketing happening with AMD slides and so on*?

AMD Analyst Day - Barcelona, Fusion and Beyond - PC Perspective

AMD Analyst Day - Barcelona, Fusion and Beyond AMD is currently holding an analyst day in which they talk about the company's current status as well as the

pcper.com

" + Title + "

2007-ish.

Compared to 2006:

AMD Analyst Day Platform Announcements

www.anandtech.com

Comparable to Torrenza's Package Integration/Core Integration(Fusion being Core Integration, but maybe with R--- product chiplet fusion will be a thing).

 

[mikegg]

Just a reminder with desktop and laptops combined the ASP for the entire computer industry, averaging celeron laptops with gaming computers and ultrabooks, all computers sold the mean is about $630 per computer sold. (It dip below $600 a few years ago, but has increased but at the same time total units sells decreased in the same timeframe so what happen was less computers were being sold but better computers.)

I bring up average ASP for the industry for you made the claim apple is going to gain 50% market share and not a more reasonable number like 20% or even 40%.

To achieve the goal you laid out then apple needs much cheaper computers, they will need computers that are cheaper than $600 to gain 50% marketshare.

That's good to know.

The M1 Mac Mini is selling for $630 on Black Friday. If Apple makes a $700 laptop, they'll likely drop the price of the Mac Mini to $600 permanently. That's below the ASP.

And Apple can achieve 50% by owning more of the high-end market, and less of the bottom of the barrel market. They've done the same for the phone, tablet, watch, and wireless headset markets with the same strategy. They own 50%+ in each of those markets in the U.S.

 

[mikegg]

You're exactly right. At a certain point it isn't worth it to try to go after more market share. If they made $600 on each of those $1000 MacBooks, dropping the price $100 requires them to sell at least 20% more MacBooks than the usually do just to make the same profit. That's certainly possible, but once you start pushing past that price level, the numbers start to make less and less sense, particularly considering that Apple would need to invest a lot of money into the additional infrastructure just to produce 20% more computers since they do some precision work on their bodies and likely have pretty strict tolerance levels as opposed to the companies that are just dumping components in cheap plastic that barely fits together.

They can build towards doing something like that slowly without running into too much trouble. However, once they start needing to produce something like three times as many computers they're going to need to spend a lot of money to have factories built to produce what they want. They certainly have the money to do that, but it's a lot of trouble to go through just to break even, never mind the initial investment. Granted the calculation is simplistic and doesn't consider fixed cost, but it's easier to look at it from the other direction. Getting to 50% market share would require Apple to sell around six to seven times as many computers as they do now.

You forgot one thing in your calculation: App Store revenue and service revenue.

Life time revenue of the customer is what you need to calculate, not hardware revenue. Wall Street is all about subscriptions now, not one-time sales. And Apple has signaled that subscriptions and services is what they will focus on.

Apple has an advantage over Dell/HP/Lenovo in that they sell additional services and subscriptions to their customers over time. iCloud, Apple Music, Apple TV, Apple News+, Apple Fitness+, Apple Arcade, a cut of the App Store revenue, selling Final Cut Pro, etc. Apple has more ways than Dell/HP/Lenovo in extracting more money from customers post-purchase.

And if there is one thing Apple is good at, it's mass manufacturing. AMD/Nvidia can't keep anything in stock. Meanwhile, more popular devices like M1 Macs can be had anywhere. Give credit to Tim Cook, who's background is in supply chain.

 

[mikegg]

You're sounding more and more ridiculous.

1) $900 is not $700.
2) Apple computers go on sale EVERY FRICKIN' YEAR on Black Friday/Cyber Monday. This is not new.
3) The list price is still $1000. $1000 is 43% more than $700.

Come back when you actually have something real to argue with.


Whatever you're smokin', I want some!

1. Yes, $900 is not $700. But I didn't say it'll happen now, right?
2. Yes, doesn't change that Apple already has the ability to drop the M1 Air to $900 and the M1 Mini to $630.
3. iPhone 12 Mini is 75% more than the iPhone SE.

Your mindset is still stuck in 2010-2014.

$270 iPad? No way.
$230 new-gen Apple Watch? What are you, crazy?
$100 Homepod? Impossible.
Apple dipping into the low-midend phone market? No way, Apple is only luxury.

Apple has gone into every the low mid-end segment in every market that they're in. But you're so sure that Apple won't for Macs?

And no, I'm not saying Apple is going to compete with $400 Costco junk laptops. They never will.

But your counter points are more and more ridiculous. Actually try to project things into the future and look at trends, instead of getting stuck in 2014.

Look at the strategy Apple has employed. Listen in to their earnings calls. Look at their market share in other markets. It doesn't take a genius to know that a Macbook SE is coming in the future.

Press Center - Apple to Start Mass Producing Self-Designed Mac SoC, Projected to Cost under US$100, in 1H21, Says TrendForce | TrendForce - Market research, price trend of DRAM, NAND Flash, LEDs, TFT-LCD and green energy, PV

According to the latest investigations by TrendForce, Apple officially unveiled its self-designed Mac SoC processors based on the ARM architecture in June, with the market release of the first Mac device with Apple Silicon (which broadly refers to chips designed by Apple) scheduled to take place...

www.trendforce.com

the production cost of a Mac processor made with TSMC’s 5nm node is currently estimated under US$100, which is considerably more cost-effective compared to the 10nm Intel Core i3 processors, priced around US$200 to US$300 on the market.

Apple's below $100 M1 is faster than Intel's' $300 mobile CPUs.

They've never done an affordable Macbook because they couldn't due to the price floor of Intel chips. Anything cheaper than Intel's i3 would be unacceptable to Apple in terms of user experience. This is no longer a problem for Apple.

 

[Roland00Address]

The problem with apple's making cheap laptops that are $600 is apple does not want to cannibalize sales and thus it does artificial market segmentation where they try to make multiple categories, where they have the best device in that category, but not a "mainstream multi use device."

Instead of selling a device that is "abc" they rather sell you device "A"+"B"+"C"

It is a different model for they can always sell you 1 more device, and if you allow me to play on this "one more thing"

(someone who is more familiar with the history will dig up 2001's Job's Digital Hub strategy and then point out the 2 pivot points since then and how this is no longer a digital hub strategy but something else. 2011 Jobs at his last cloud says iCloud "launching today" will replace the former digital hub, and there has to be a pivot or two since then.)

----

Now $600 or even $500 mac minis are possible but I doubt $600 macbook airs even if I see $800 macbook airs as possible (and $800 will always have 1 or 2 fatal flaws like needing more ram or storage, something to make you pay $200 more at least.)

 

[mikegg]

Now $600 or even $500 mac minis are possible but I doubt $600 macbook airs even if I see $800 macbook airs as possible (and $800 will always have 1 or 2 fatal flaws like needing more ram or storage, something to make you pay $200 more at least.)

I don't think $600 Macbook Airs are possible. I think Apple will do a $700 Macbook SE. But they could also do $725 or $750. The exact price doesn't matter too much, but it needs to be below $800 to have any impact in capturing major marketshare.

M1 Mac Minis are already on sale for $630.

 

[amrnuke]

That's good to know.

The M1 Mac Mini is selling for $630 on Black Friday. If Apple makes a $700 laptop, they'll likely drop the price of the Mac Mini to $600 permanently. That's below the ASP.

And Apple can achieve 50% by owning more of the high-end market, and less of the bottom of the barrel market. They've done the same for the phone, tablet, watch, and wireless headset markets with the same strategy. They own 50%+ in each of those markets in the U.S.

If you've found a good way to convince curmudgeonly executive board rooms and government agencies worldwide to abandon Windows and retool their IT/support infrastructure, and to do it within 5 years, and at a higher cost, then you may have a chance at being within 30-40% of the right answer on Apple's marketshare in 5 years.

 

[mikegg]

Just... wow.

If you've found a good way to convince curmudgeonly executive board rooms and government agencies worldwide to abandon Windows and restructure their whole enterprise's IT department, and to do it within 5 years, then you may have a chance at being close to correct.

It's not going to happen overnight. Some places will never switch.

And I've already said in this forum that it wouldn't surprise me if they actually got to 50% in 5 years, but I expect it to happen within 10 years.

By the way, many enterprises will switch. They just couldn't justify $1200 Macbooks for their $30,000/year employees.

Macs at the enterprise level costs less to maintain and has better satisfaction levels:

Total cost of ownership: Mac versus PC in the enterprise. Apple TCO

If the upfront cost is more, how can Mac be less expensive than PC? We examine the numbers and put the Mac versus PC in the enterprise myths to bed.

www.jamf.com

Apple is a top three PC vendor for enterprise, report claims

When it comes to corporate notebook purchasing, Macs are just as likely a choice as HP, with Dell still ahead, according to 451 Research.

www.computerworld.com

"In fact, IBM found they saved between $273 - $543 per Mac they deployed compared to PCs."


Any worthwhile enterprise software has moved to the cloud (browser-based). What's left is internal software written for Windows or niche x86 software. It's much easier for enterprises to switch now than say in the 90s or 2000s or 2010s.

 

[LightningZ71]

It's not going to happen overnight. Some places will never switch.

And I've already said in this forum that it wouldn't surprise me if they actually got to 50% in 5 years, but I expect it to happen within 10 years.

By the way, many enterprises will switch. They just couldn't justify $1200 Macbooks for their $30,000/year employees.

Macs at the enterprise level costs less to maintain and has better satisfaction levels:

Total cost of ownership: Mac versus PC in the enterprise. Apple TCO

If the upfront cost is more, how can Mac be less expensive than PC? We examine the numbers and put the Mac versus PC in the enterprise myths to bed.

www.jamf.com

Apple is a top three PC vendor for enterprise, report claims

When it comes to corporate notebook purchasing, Macs are just as likely a choice as HP, with Dell still ahead, according to 451 Research.

www.computerworld.com

This last point I will easily support! In the last decade, I have personally watched several organizations switch from traditional tower PCs to uSFF PCs that have just one or two USB connected devices. 99% of what they do is either browser based, or written in Java. The PCs are essentially just throw away boxes. Some didn't even bother with extended warranties and just bought a few extra units instead. I am certain that two of the organizations could switch to Mac Minis right now without skipping a beat.

"In fact, IBM found they saved between $273 - $543 per Mac they deployed compared to PCs."


Any worthwhile enterprise software has moved to the cloud (browser-based). What's left is internal software written for Windows or niche x86 software. It's much easier for enterprises to switch now than say in the 90s or 2000s or 2010s.

 

[amrnuke]

It's not going to happen overnight. Some places will never switch.

And I've already said in this forum that it wouldn't surprise me if they actually got to 50% in 5 years, but I expect it to happen within 10 years.

By the way, many enterprises will switch. They just couldn't justify $1200 Macbooks for their $30,000/year employees. But what if they can get $600 Macbook SEs at volume pricing?

Macs at the enterprise level costs less to maintain and has better satisfaction levels:

Total cost of ownership: Mac versus PC in the enterprise. Apple TCO

If the upfront cost is more, how can Mac be less expensive than PC? We examine the numbers and put the Mac versus PC in the enterprise myths to bed.

www.jamf.com

Apple is a top three PC vendor for enterprise, report claims

When it comes to corporate notebook purchasing, Macs are just as likely a choice as HP, with Dell still ahead, according to 451 Research.

www.computerworld.com

"In fact, IBM found they saved between $273 - $543 per Mac they deployed compared to PCs."

Now we're getting $600 Macbooks. Wow! What a shift in Apple's vision! That's cheaper than the iPhone mini!

As for the JAMF study, as someone who has been doing actual research for the better part of 15+ years, not controlling for the variables in that study is absurd.

Let me explain why the study has to be taken with a grain of salt: People who choose Mac (as they did in the Mac@IBM program) are far different from those who have no preference or who declined to be issued a Mac. What they should have done is issued Mac or PC randomly to a subset of employees and then measured the differences. All the other results are essentially useless.

For example: I have an iPhone at home and bought it because I like it more than Androids. If my employer gave me a choice between a Galaxy and an iPhone and I picked an iPhone, would it be surprising if I told the surveyors that I like my own choice of phone, or that it made me more productive? I also am far less likely to call IT than someone with a Droid at home, who got issued an iPhone. Both subsets are far less likely to call tech support than someone who uses only a landline at home.

Of course someone who chooses Mac is going to require less tech support than someone who takes whatever is given to them - because the person choosing a Mac is more likely to have had prior experience with it, and probably good experience, hence their choice. And when you ask someone if they liked their choice, most people aren't going to say, "Nah, I made a bad choice."

He said it poorly at the end: "I don't know if better employees want Macs, or if giving Macs to employees makes them better. You gotta be careful about cause and effect."

He left out the most important piece -- We didn't construct this study in a manner that can properly draw any conclusions as to why people who selected Macs tended to perform better. Could it be because of halo effects? Because "creatives" tend to gravitate toward the "creative"-oriented Mac? Because people who have technology preferences tend to be a different type of employee? No, we'll just give an either-or of two improbable options: 1) if you don't want a Mac we know you're not as good an employee, or 2) a Mac is a panacea for all enterprise woes.

BTW, they must not be terribly convinced of the results he presented. Even in one of the most forward-thinking companies with all the infrastructure and experience and high level intellect and adaptability you could want, with the first deployment of Macs starting in 2015 (hey, 5 years ago!), they still use 60% Windows and 10% Linux and only 30% Macs.

Apple is not getting 50% marketshare in 2025 or in 2030.

 

[jeanlain]

I see that this thread is no longer about the M1 or Apple SoCs, but about Apple in general.

 

[amrnuke]

I see that this thread is no longer about the M1 or Apple SoCs, but about Apple in general.

So back to the M1, I found LTT's latest review of the Mac Mini interesting.

Very interesting the H.265 vs H.264 hardware results. This just goes along with their extreme fine-tuning with the mobile / tablet experience. The most important things for most users of the MBA, MBP, and Mac mini is going to be responsiveness and multimedia experience.

Another interesting thing he mentions is the implementation of Rosetta. I'm not sure what exactly is meant by x86 consistency model and how it affects direct execution of ARM-like x86 instructions or if it enables direct execution of certain x86 instructions. Anyone care to take a layman's jab?

 

[jeanlain]

Very interesting the H.265 vs H.264 hardware results. This just goes along with their extreme fine-tuning with the mobile / tablet experience. The most important things for most users of the MBA, MBP, and Mac mini is going to be responsiveness and multimedia experience.

For x265 at least, the ARM executable has only few neon optimisations, and Apple's contributions have not yet been implemented.

 

[Viper GTS]

So back to the M1, I found LTT's latest review of the Mac Mini interesting.

Very interesting the H.265 vs H.264 hardware results. This just goes along with their extreme fine-tuning with the mobile / tablet experience. The most important things for most users of the MBA, MBP, and Mac mini is going to be responsiveness and multimedia experience.

Another interesting thing he mentions is the implementation of Rosetta. I'm not sure what exactly is meant by x86 consistency model and how it affects direct execution of ARM-like x86 instructions or if it enables direct execution of certain x86 instructions. Anyone care to take a layman's jab?

He is referring to this:

https://twitter.com/x/status/1331736203402547201

I'm sure someone has done or will do a more in depth technical explanation but basically Apple is doing x86 style memory operations in hardware rather than emulating them They can do either method depending on what the application needs which gives them a massive performance advantage.

Viper GTS

 

[Heartbreaker]

He is referring to this:

I'm sure someone has done or will do a more in depth technical explanation but basically Apple is doing x86 style memory operations in hardware rather than emulating them They can do either method depending on what the application needs which gives them a massive performance advantage.

That memory ordering thing, is referring to Big Endian(Moto 68K) vs Little Endian (x86).

ARM is Bi-Endian (it goes both ways), as was PowerPC, so this isn't new, nor it is in Apple unique feature.

 

[moinmoin]

That's pretty clever.

 

[Hitman928]

That memory ordering thing, is referring to Big Endian(Moto 68K) vs Little Endian (x86).

ARM is Bi-Endian (it goes both ways), as was PowerPC, so this isn't new, nor it is in Apple unique feature.

No, he is referring to x86 using TSO for memory consistency, it has to do with read/write conflicts for multi-core architectures. He is saying that M1 can switch between TSO when running Rosetta and whatever it uses natively. I'm less familiar with ARM but I believe they use a relaxed memory consistency model more similar to POWER than x86. I have no idea if this is true or not but that is what he is claiming. If true, I also wonder if Intel tries to sue depending on how it is being implemented in the M1 architecture as they have been quick to sue other companies that have tried to implement similar hardware features though I think that was more to do with the instruction set itself.

 

[jeanlain]

(ignore)

 

[Schmide]

ARM is Bi-Endian (it goes both ways), as was PowerPC, so this isn't new, nor it is in Apple unique feature.

Before this gets out of hand. Little endian for life. Personal opinion.

and most user space these days is little endian. Except for some low power micro controllers (AVR and the like) and IBM Power.

Though openRISC is big endian.

Regardless. If you've been coding on ARM in the past 10 years. The OS has used little endian.