The case of an Ultra chip using two M2 Pro chips

[Rickroller]

Another thing to consider is how the Max was designed to produce Ultras. To my understanding the Pro may not be setup in a way to create Pro-based Ultras.

Any supply chain person would limit their SKUs to the most saleable ones so inventory remains lean. Your use case does not appear to be in a volume that Apple sees worthwhile to the bottomline.

Die shrinks do allow for further reduced power consumption. See the iPads & iPhones of late. Battery life extended for the MBA M1 vs to Intel Mac.

Not sure the M2 Pro can’t produce an Ultra version. I’m just spitballing really. I wonder if you were were aware that M1 Ultra was possible before Apple showed it off. As for product volume, there actually isn’t a saturated marker for Apple silicon hardware. The Ultra is currently only available in the Studio, and obviously not a cheap computer by any means. The only other Apple product that can house it is a Mac Pro which is the real limitation if we take cost out of it.

Yes you are correct that die shrinks reduce power consumption, but you do realize that the iPhone A series has been doing that for pretty long time now and it’s still a 5w part.

 

[Rickroller]

Cheaper based on silicon wafer surface area (material) but not leading edge node fab equipment.

So why would they rush to the next node which is by all accounts far more expensive than pushing the current one a little bit further…?

 

[sam_dean]

So why would they rush to the next node which is by all accounts far more expensive than pushing the current one a little bit further…?

For the reason I bullet pointed above and product positioning.

Leading edge node provides them with an advantage over older ones.

You can stay stagnant on a node only if you have a monopoly.

Example would be Motorola and the Power PC chip. Apple had to ditch them for Intel by 2006 because Motorola would not or could not move to a leading edge node. Steve was adamant about the move because by that time laptops were 80% and desktops 20% of the PC market. Steve famously said he wanted to sell us Powerbook G5 laptops but could not do so because they ran too hot and draw too much power.

When Apple made that move it made Intel into a monopoly.

Fastforward to 2014-2020 and Intel ceased any actual effort to move to the next node to increase shareholders value.

Apple realized they had resources to circumvent Intel at their disposal. They were shipping approx quarter billion iPhone/iPad chips per year at the leading node. That's the number of Intel chips when Apple was still their customer.

iOS/iPadOS was based on macOS so it isn't that hard to make the move as they've always been hardware agnostic,

Adding nearly 29 million Mac chips is small potatoes by comparison and helps increase their purchasing position with TSMC and other relevant suppliers.

It is all about the supply chain. Master it and you'll make $AAPL money.

Your ideas nice and well thought out but supply chain and chip design limits your idea.

 

[Rickroller]

For the reason I bullet pointed above and product positioning.

Leading edge node provides them with an advantage over older ones.

You can stay stagnant on a node only if you have a monopoly.

Example would be Motorola and the Power PC chip. Apple had to ditch them for Intel by 2006 because Motorola would not or could not move to a leading edge node. Steve was adamant about the move because by that time laptops were 80% and desktops 20% of the PC market. Steve famously said he wanted to sell us Powerbook G5 laptops but could not do so because they ran too hot and draw too much power.

When Apple made that move it made Intel into a monopoly.

Fastforward to 2014-2020 and Intel ceased any actual effort to move to the next node to increase shareholders value.

Apple realized they had resources to circumvent Intel at their disposal. They were shipping approx quarter billion iPhone/iPad chips per year at the leading node. That's the number of Intel chips when Apple was still their customer.

iOS/iPadOS was based on macOS so it isn't that hard to make the move as they've always been hardware agnostic,

Adding nearly 29 million Mac chips is small potatoes by comparison and helps increase their purchasing position with TSMC and other relevant suppliers.

It is all about the supply chain. Master it and you'll make $AAPL money.

Your ideas nice and well thought out but supply chain and chip design limits your idea.

The thing is I only saw this as an extra on top chip that could be useful for segmentation for some of their products. I know that it might seem silly to suggest this, but Apple currently do have somewhat of a monopoly when it comes to native Mac OS systems.

I think if Apple want to attack the lucrative engineering workstation market, they would want to have more than one option that can deliver performance at that level. I can’t imagine they’re not doing something on that from in regards to software that is specifically built for Apple silicon, as that is another way to create non competitive environment for the PC makers.

Intel is actually going to be releasing competitors to AMD Threadripper, and those go for $6000 for the 5990 skew. I doubt they’d do that if it wasn’t worth the effort financially. So perhaps the market isn’t as small for HPC as one would think…

I would love to see if there are any actual cost breakdowns for Apple chips compared to what you get from the other guys…like what a full fat Ultra would cost if they sold it as a part instead of as a part of a complete product.

 

[sam_dean]

The thing is I only saw this as an extra on top chip that could be useful for segmentation for some of their products. I know that it might seem silly to suggest this, but Apple currently do have somewhat of a monopoly when it comes to native Mac OS systems.

I think if Apple want to attack the lucrative engineering workstation market, they would want to have more than one option that can deliver performance at that level. I can’t imagine they’re not doing something on that from in regards to software that is specifically built for Apple silicon, as that is another way to create non competitive environment for the PC makers.

Intel is actually going to be releasing competitors to AMD Threadripper, and those go for $6000 for the 5990 skew. I doubt they’d do that if it wasn’t worth the effort financially. So perhaps the market isn’t as small for HPC as one would think…

I would love to see if there are any actual cost breakdowns for Apple chips compared to what you get from the other guys…like what a full fat Ultra would cost if they sold it as a part instead of as a part of a complete product.

This video may cause you to revise some of your assumptions.

Your use case is very legitimate but not at the units shipped that Apple see as profitable.

Apple does not sell chips/parts for other manufacturers. Microsoft does not offer a version of winOS that would run on it. Linux? Linux you say? Linux on the desktop? 🤣 That's been a running joke since the 90s.

Apple has always monopolized macOS. It isn't offered to other PC maker.

"Lucrative" relative to what? Apple improved their iOS gaming business because it was a better use of their money. Apple Pay has more of a runway than anything else.

Engineering workstations may have a great margin but the volume isn't worth their undevided focus.

Apple would inevitably encounter the same concerns of "not upgradeable" that the engineering workstation market would demand.

 

[Rickroller]

This video may cause you to revise some of your assumptions.

Your use case is very legitimate but not at the units shipped that Apple see as profitable.

Apple does not sell chips/parts for other manufacturers. Microsoft does not offer a version of winOS that would run on it. Linux? Linux you say? Linux on the desktop? 🤣 That's been a running joke since the 90s.

Apple has always monopolized macOS. It isn't offered to other PC maker.

"Lucrative" relative to what? Apple improved their iOS gaming business because it was a better use of their money. Apple Pay has more of a runway than anything else.

Engineering workstations may have a great margin but the volume isn't worth their undevided focus.

Apple would inevitably encounter the same concerns of "not upgradeable" that the engineering workstation market would demand.

Yes thanks for this…the stuff that they talk about happened way before Su got there…timeline isn’t possible unless she has a Time Machine!

Unlike Intel and AMD, Apple makes profits from complete hardware not just the chips powering them. So the appeal of their products as consumer devices is far more relevant in their case. They don’t need to sell more than the other guys to be profitable…people couldn’t believe iPhones were $600 and now are over $1000. What they do need to have is something that the other guys can’t compete with tho.

 

[sam_dean]

Yes thanks for this…the stuff that they talk about happened way before Su got there…timeline isn’t possible unless she has a Time Machine!

Unlike Intel and AMD, Apple makes profits from complete hardware not just the chips powering them. So the appeal of their products as consumer devices is far more relevant in their case. They don’t need to sell more than the other guys to be profitable…people couldn’t believe iPhones were $600 and now are over $1000. What they do need to have is something that the other guys can’t compete with tho.

Apple does not need to be present in all markets at all times.

They let other players cater to markets where in their technology would be most effective.

 

[leman]

why is there an expectation that M3 is going to be available so soon…? I would guess that iPhone takes priority for any new chips considering how many units they sell of those vs MacBooks…

Not as much an expectation as speculation. As to the reasons, the main one is the Mac Pro. I just have a feeling that we might have to see some surprises at this WWDC. I think the real high-performance Apple Silicon was always planned for 3nm.

iPhone is less of a priority probably. Initial 3nm is going to be quite expensive and A16 is already ahead of the rest of the market by a year or two. Makes more sense to reserve that initial capacity towards the high-end.

Personally I think that Apple isn’t really going for stopgap solutions as it’s not like their chips are underwhelming in an Apple silicon market that’s barely over two years old. The whole marketing around M2 is still in comparison to Intel systems, so they must have data telling them there are still a large pool of holdouts to convert.

Looking at the new MacBook Pros just released, they seem to have been on track for a 12 month update if something unexpected hadn’t happened.

There is a lot of evidence that Apples execution has been massively delayed. I believe M1 Pro/Max/Ultra were supposed to launch in summer 2021 and M2 in fall 2021. 3nm chips were planned to be on the market by now.

 

[sam_dean]

Not as much an expectation as speculation. As to the reasons, the main one is the Mac Pro. I just have a feeling that we might have to see some surprises at this WWDC. I think the real high-performance Apple Silicon was always planned for 3nm.

iPhone is less of a priority probably. Initial 3nm is going to be quite expensive and A16 is already ahead of the rest of the market by a year or two. Makes more sense to reserve that initial capacity towards the high-end.




There is a lot of evidence that Apples execution has been massively delayed. I believe M1 Pro/Max/Ultra were supposed to launch in summer 2021 and M2 in fall 2021. 3nm chips were planned to be on the market by now.

It is severely unlikely that M3 will get 3nm ahead of the iPhone chips as Apple makes more money from phones than from the Mac.

3nm likely timeline

- Sep 2023: iPhone chip... likely iPhone 15 Pro & iPhone 15 Pro Max
- Oct/Nov 2023: M3 chip... likely MBA & MBP 13"... possibly Jan 2024

 

[sam_dean]

BTW your title's a bit confusing.

Consider editing it to: "The case of an Ultra chip using two M2 Pro chips"

 

[Rickroller]

BTW your title's a bit confusing.

Consider editing it to: "The case of an Ultra chip using two M2 Pro chips"

Actually I will do that as you’re probably right that it is a little confusing.

 

[Rickroller]

Not as much an expectation as speculation. As to the reasons, the main one is the Mac Pro. I just have a feeling that we might have to see some surprises at this WWDC. I think the real high-performance Apple Silicon was always planned for 3nm.

iPhone is less of a priority probably. Initial 3nm is going to be quite expensive and A16 is already ahead of the rest of the market by a year or two. Makes more sense to reserve that initial capacity towards the high-end.




There is a lot of evidence that Apples execution has been massively delayed. I believe M1 Pro/Max/Ultra were supposed to launch in summer 2021 and M2 in fall 2021. 3nm chips were planned to be on the market by now.

Ha! I was thinking the same thing in terms of surprises for the Mac Pro as well, but I’m not sure a node shrink is what one would consider a surprise as such…

To really pull off something unexpected would mean presenting something that blows the doors off what is currently seen as possible. I’m gonna go out on a limb and speculate that Apple is going to pull off the first 3D chip packaged from a stacked set of Ultras. As in two Ultra chips stacked on top of each other somehow.

This would mean that technically they don’t have an M2 Extreme per se, as that was supposedly too difficult or expensive to produce, but maybe they were late enough to get the 3D design goal that the puck seems to actually be skating to.

 

[senttoschool]

Not as much an expectation as speculation. As to the reasons, the main one is the Mac Pro. I just have a feeling that we might have to see some surprises at this WWDC. I think the real high-performance Apple Silicon was always planned for 3nm.

iPhone is less of a priority probably. Initial 3nm is going to be quite expensive and A16 is already ahead of the rest of the market by a year or two. Makes more sense to reserve that initial capacity towards the high-end.




There is a lot of evidence that Apples execution has been massively delayed. I believe M1 Pro/Max/Ultra were supposed to launch in summer 2021 and M2 in fall 2021. 3nm chips were planned to be on the market by now.

iPhone will always eat first when it comes to wafers. No question about it. There should be enough 3nm wafers to go around since Apple is the only customer for 3N.

I too think that M1 Pro/Max was delayed, M2 was delayed, and M2 Pro/Max were also delayed. M1 Pro/Max was suppose to launch in Summer 2021 like you said. There were plenty of reputable rumors. M2 not launching until Summer 2022 was a surprise to me since it was just based on the A16. Rumors were that it was suppose to launch in Spring 2022 but the Macbook Air M2 redesign was what caused the delay. Finally, there is evidence that suggest Apple wanted to launch M2 Pro/Max in October 2022.

I'm guessing that Apple didn't hold an official event to announce the M2 Pro/Max because we're only 4 months away from WWDC and they could be launching the M3 at that time. They probably just felt like it was too late to be heavily promoting an M2 based SoC when M3 is right around the corner. Gurman did say that M2 is likely to be a short lived generation and I believe that.

Just speculation but I think the ideal schedule for Apple is this:

• Fall: New iPhone SoC & new base M chip. Capture holiday shopping with new models for the most popular Mac: Macbook Air.
• Spring: Pro and Max versions. Gives a few months extra time because Pro/Max are more complicated SoCs.
• WWDC: Ultra and Extreme versions. Makes sense to announce them at WWDC because these SoCs are more tailored to creators and professionals.

And repeat once a year.

 

[Appletoni]

I think there is a pretty good chance that there will be a second Ultra version for the M2 family of chips. The Pro chip isn’t a cut down version of the Max chip this time around, it’s designed as a unique full unit, and is probably capable of being used in an Ultra Fusion assembly when scaled up. The upcoming Mac Pro would benefit from having more than one option powering its chassis, and the more balanced M2 Pro could do the trick.

there is also another reason to add this chip to the family, and that’s to allow Apple to add an Ultra addition to the MacBook lineup. When you look at what Inten and AMD are planning to release as soon as next month for the X86 laptop market, it would be an opportunistic marketing coup that would at least place some competitive CPU performance pressure where they may have thought there was none.

As Apple generally designs for a specific thermal power envelope, I think a possible M2 Pro Ultra should be fine in the 16 inch Pro chassis. Doubling the wattage of the Pro chip to 120W is of course more than the M2 Max 90W limit, but we already have the Max chip in the 14 inch MacBook Pro, and that also has a relatively
small chassis and battery for the Chip it’s running.

Also adding this less beastly Ultra would help to close the gap in the Mac Studio lineup. At the moment the base model starts at $1999 for the Max and Jumps to the M1 Ultra at $3999, which isn’t ideal. Another good thing about the Pro Ultra is that for some users the extra cores are more important than the extra GPU cores.

The M2 Pro Ultra

24 CPU cores
38 GPU cores
Double neural
double double all the codecs
and 144GB RAM @ 600 GB/s Bandwidth

Even the Apple fanboys proved months ago that it can handle the best high-end Threadripper CPU.

There is no problem with the heat.
How many games do you know that absolutely need all GPU cores?
How many games do you know that need all 20 CPU cores at the same time? Most of them need 1-2 CPU cores.
How many games do you know that use 128 GB RAM at the limit?
How many games do you know that cause massive SSD accesses?
...
How many games do you know that do all this at the same time?
There will be no heat issues with games.

Only scientists and chess players (and a handful of exceptions) could run these devices at the limit.

That being said, if needed, heat is lowered by increasing fan speed.
Or by slower clocking cores.
Or by a mixture of everything.
And even then, the MacBook with M1 ULTRA will still be significantly faster than an M1 MAX.
M2 ULTRA... M2 MAX.

For most buyers of an Apple MacBook Pro 16-inch M1/M2 ULTRA, there will never be anything even close to heat issues.

Additionally, Apple can easily:
-use a high-end thermal compound (easy).
-install better heatsinks (easy)
-improve the rotation speed of the fans (easy for Apple)
-improve fan noise (easy for Apple)
-improve the airflow of the fans (easy for Apple)
-use a better MacBook Pro case (medium difficulty for Apple)
-offer larger MacBook Pro enclosures in 18-inch and in 20-inch (easy for Apple)

 

[Appletoni]

Very unlikely. It's not just about thermals, but also the product positioning. Apple definitely wouldn't want too much overlap between different segments. Ultra chip would be too overpowered for a laptop.

It depends on the users what they need the MacBooks and their performance for.
Apple MacBook with M1 Extreme chip would be too underpowered for many people and me in a laptop.

 

[sam_dean]

At $3999 both the 2021 & 2023 Mac Studio Ultra 5nm should have 128GB RAM & 2TB SSD.

That config should be good until year 2033 when sub-1nm chips are found in Macs & iPhones.

 

[Appletoni]

I think what prevents the 5nm M1/M2 Ultra being in a MBP 16" is the

- die shrink
- power consumption
- thermals
- price points

Ultra doubles the price of a Mac Studio. Is there a large enough market for a laptop of that price point?

3nm M3 Ultra in a MBP 16" may occur in mid 2024 to early 2025.

As for Mac chip with mostly CPU cores & lesser GPU cores or mostly GPU cores and lesser CPU cores it appears that Apple is focusing on volume use cases and not unique ones.

Unique ones like Wolfram Mathematica users who want as much CPU cores as possible with as few GPU cores. Odds are scientists do not outnumber gamers, gaming app developers or 8K video editors. As such most of the SoCs are for volume use case.

Die shrink is not a (problem).
More power consumption means more heat. But this is not a problem either, as long as the power consumption does not become too large.
The price is not a problem.
If the device is worth the price, or if you absolutely need it, or if you absolutely want it, you pay the price.

Yes the market for a MacBook Pro with an M1 ULTRA / M2 ULTRA / M3 ULTRA chip is big enough.
If buyers get almost all the ports on the MacBook that the Mac Studio has and everything else is high-end, then the MacBook has the advantage of another screen and the ability to do everything possible on the go or in different places at home.

Now of course it could be that 99.99% of all customers will buy the MacBook with ULTRA chip instead of the Mac Studio with ULTRA chip.

 

[Appletoni]

i think the 16 inch MacBook Pro should be able to handle a 120w chip, and I’m not sure that there are a lot of workloads that would pin both CPU and GPU to the limit for extended periods of time out side of benchmarking.

Pricing is something to consider of course, but what is a halo product such as super-powered MacBook Ultra worth…? Changing the logic board and beefing up cooling is doable I would think, and adding $1000 would sting for sure, but you can’t buy a Ferrari for Camry prices either.

A Mac Studio with this chip should be a reasonable tier at $2999…? Why is a user who only needs the CPU from the Ultra forced to pay for the extra GPU…? I bet the number in this boat isn’t insignificant when you look at the price jumps to the higher tier stuff.

2024 is when I think the next series of M chips arrive, but that’s a while from now. The die shrink is used to add cores, not reduce the the package power. As in the M3 chips on 3nm will still mean the M3 Pro will be a 60w part, and the M3 Max will be a 90w part, so still to power hungry in a MacBook in an Ultra with 180w.

A MacBook Pro is sold to customers who use them to make money. Especially the 14 and 16 inch models. It’s a very profitable section of the market, and unsurprisingly they built their silicon to cater directly to the specific workloads these user target. certainly gaming is not part of the equation.

Gaming is always a target.
My M1 MAX maxed out is great for macOS / iPadOS / Windows 11 ARM / Linux gaming.

 

[Appletoni]

Ha! I was thinking the same thing in terms of surprises for the Mac Pro as well, but I’m not sure a node shrink is what one would consider a surprise as such…

To really pull off something unexpected would mean presenting something that blows the doors off what is currently seen as possible. I’m gonna go out on a limb and speculate that Apple is going to pull off the first 3D chip packaged from a stacked set of Ultras. As in two Ultra chips stacked on top of each other somehow.

This would mean that technically they don’t have an M2 Extreme per se, as that was supposedly too difficult or expensive to produce, but maybe they were late enough to get the 3D design goal that the puck seems to actually be skating to.

I think the same about the first 3D chip packaged. Hope to see it soon.
With the M2 Extreme, Apple is unsure whether there will be enough buyers.
That is why the M2 Extreme is paused.
Everything else is not a problem.

 

[Appletoni]

iPhone will always eat first when it comes to wafers. No question about it. There should be enough 3nm wafers to go around since Apple is the only customer for 3N.

I too think that M1 Pro/Max was delayed, M2 was delayed, and M2 Pro/Max were also delayed. M1 Pro/Max was suppose to launch in Summer 2021 like you said. There were plenty of reputable rumors. M2 not launching until Summer 2022 was a surprise to me since it was just based on the A16. Rumors were that it was suppose to launch in Spring 2022 but the Macbook Air M2 redesign was what caused the delay. Finally, there is evidence that suggest Apple wanted to launch M2 Pro/Max in October 2022.

I'm guessing that Apple didn't hold an official event to announce the M2 Pro/Max because we're only 4 months away from WWDC and they could be launching the M3 at that time. They probably just felt like it was too late to be heavily promoting an M2 based SoC when M3 is right around the corner. Gurman did say that M2 is likely to be a short lived generation and I believe that.

Just speculation but I think the ideal schedule for Apple is this:

• Fall: New iPhone SoC & new base M chip. Capture holiday shopping with new models for the most popular Mac: Macbook Air.
• Spring: Pro and Max versions. Gives a few months extra time because Pro/Max are more complicated SoCs.
• WWDC: Ultra and Extreme versions. Makes sense to announce them at WWDC because these SoCs are more tailored to creators and professionals.

And repeat once a year.

Everything was delayed several times.
I was already planning to buy the MacBook Pro 20-inch with doubled EXTREME chip in 2019.
And 512 GB RAM.
And 32 TB SSD.
For Chess ;-)

 

[Rickroller]

Gaming is always a target.
My M1 MAX maxed out is great for macOS / iPadOS / Windows 11 ARM / Linux gaming.

That’s where we differ in thinking…I know that Apple considers gaming as an important part of the market, but I’m not sure the high end gaming market is part of it. The strange things you hear on this forum referring to to Apple not being able to compete with Nvidia 4090s. Those are $1600 a pop!

How big is the audience for gaming pc upwards of $5K…? Especially if they want to get margins of 40%…? If they do plan on a gaming focused device, it will probably be a console type box like maybe an Apple TV Pro or Max. Maybe the strategy is gonna be planned around AR/VR.

 

[sunny5]

The max power consumption for M1 Ultra is 60W + 120W. M2 version will consume more power so it will defeats the purpose of MBP and wont be able to charge with USB-C and MagSafe 3. Unless Apple is willing to make a such thing, I dont think it will fit for laptop series. The battery is the biggest issue which didnt really improved and advanced for several decades.

Might be a good idea if M2 Ultra MBP works at 100% of power as long as it's connected to the power source just like other PC laptops and that's the only solution. Without the power, reduce the max power consumption to 100W. Then, it might make sense.

 

[Rickroller]

The max power consumption for M1 Ultra is 60W + 120W. M2 version will consume more power so it will defeats the purpose of MBP and wont be able to charge with USB-C and MagSafe 3. Unless Apple is willing to make a such thing, I dont think it will fit for laptop series. The battery is the biggest issue which didnt really improved and advanced for several decades.

Might be a good idea if M2 Ultra MBP works at 100% of power as long as it's connected to the power source just like other PC laptops and that's the only solution. Without the power, reduce the max power consumption to 100W. Then, it might make sense.

I think my title might have confused you, I was proposing a new lower power ultra made up of two M2 Pro chips That would/should fit in a 16 inch MacBook chassis. That would mean a 120w part that is under the 140w charger currently included. The 14 inch pro can be optioned with the M2 Max which is a 90w part When it only has a 70w battery.

 

[Rickroller]

Just as an exercise, I looked up what is on the table from the competition from the other guys

Look at the watts!!!

 

[JouniS]

Just as an exercise, I looked up what is on the table from the competition from the other guys

Look at the watts!!!

Intel CPUs are one thing, but Nvidia GPUs look quite good.

Based on the number of CUDA cores, the mobile 4070 is comparable to the M2 Max GPU, which uses >50 W at ~1400 MHz. According to the numbers you quoted, the 4070 may be even more efficient. Similarly the mobile 4090 is comparable to the hypothetical M2 Ultra GPU, with quite decent power usage at lower clock rates.