Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

sunny5

macrumors 68000
Original poster
Jun 11, 2021
1,740
1,608
Pure curiosity before Mac Pro shows up later this year. Let's say a new chip's name for Mac Pro is M2 Extreme.

1. M2 Extreme will be a powerful chip but not in GPU because many workstations already can use more than 1 graphic card or can be used up to 4 or more. Mac Pro 2019 supports up to 4 graphic cards. Which means a new Mac Pro needs up to 4x M2 Extreme chips. If you are fine with just one M2 Extreme, then that's fine but how about those people who need more than just one graphic card? Are you gonna say you dont need more than 1 graphic card lol. There are reasons why many workstations support multiple GPU and you never know if Mac Pro becomes a beast in 3D market.
Adding an external GPU might be the option but Apple Silicon take advantages from unified memory and SoC so I doubt it.

2. The maximum RAM size is in question. With only one M2 Extreme, it might have up to 256gb or maybe up to 384gb based on M2's memory. Well, that's way lower than Mac Pro's maximum RAM size which is 1.5TB. Others can support up to 4TB. Even if they can use 4x M2 Extreme, it's barely 1.5TB. Yes, that's only IF they use 4 chips. And Yes, they might be able to add a lot of memory chips along with one M2 Extreme but can they really add those memories around the chip?

3. Will Mac Pro supports PCIe slots? What about upgradability and expandability? Currently, Apple is very hostile toward both area even for Mac desktops. If Apple wants Mac Pro to be another Mac Studio, what's the point? So far, I'm quite skeptical that Apple is willing to support PCIe slots and other upgradable parts.

I have no idea how AS Mac Pro will be look like but it shouldn't be another Mac Studio or Mac Pro 2013. If not, then what's the point of having Mac Pro? Mac Pro 2013 was a huge failure anyway.

We'll see after Apple announces a whole new Mac Pro...
 
  • Like
Reactions: maxoakland

trevpimp

macrumors 6502a
Apr 16, 2009
697
301
Inside A Mac Box
I'm guessing the Mac Pro will not be upgradeable with nothing other than Apple optional counterparts

They will lock in SoC counterparts that way we wouldn't have any options other than to choose from them

"In A Good Way" it leaves all other competitions from beating Mac and makes Mac Pro great for professionals and ahead of the SoC game
 

PsykX

macrumors 68030
Sep 16, 2006
2,682
3,803
Quick question, because I genuinely don't know : Who/What type of workflow uses more than 256GB of RAM?

Second question : When people go for Mac Pros, what percentage of users buy configs over 256GB ?

Everything points towards 384GB. Yes, it's a huge downgrade in terms of capacity. But I think Apple will let it go because above that is an extremely tiny market share.
 

EugW

macrumors G5
Jun 18, 2017
14,443
12,279
I had been guessing that the Mac Pro chips could have a differently designed SoC, with a different memory interface, and different naming convention. But I could be completely wrong on that of course. ;)
 

innerproduct

macrumors regular
Jun 21, 2021
208
332
There is a "huge" crowd of creatives that need 3D rendering power. These guys are both freelancers and part of larger studios. They (we) need responsiveness during the whole pipeline from concept through modelling, lookdev, lighting etc. Freelancers usually buy a bang for the buck machine and for a few years that has been a 16 core AMD (1950, 2950, 2950, 5950) coupled with dual GPUs, atm that would be dual 3060 tis -> 3090 tis. This is a config that solves all the needs for the category of user. I hoped for a long time that we could get something "close enough" on the Mac side and it was pretty close with a iMac 27" and dual eGPUS with 5700 ->6900xt cards in. Sadly that is not a thing anymore and we are left with the stinking mess that is SoC. The studio ultra has a nice CPU but the GPU is a joke. For 3D rendering it is at best Nvidia 3060 levels... So, there desperately needs to come something that is not just a doubling of GPU perf. Upcoming competition from Nvidia will have GPUs in the 100 TFLOPS range and AMD in 75 TFLOPS range. Even without counting the specialised ray tracing cores. (and these theoretical figures actually maps almost perfectly to 3d rendering performance in SW like blender, redshift and octane). So this fall a freelancer can buy a AMD 7950 and have about 200 TF of rendering power. That is a magnitude better than what the best Mac can provide even assuming we had optimised SW for the Mac as well. I mean, who would buy a machine that is at best 10% a fast as the competition and cost twice as much? It is so dumb a proposition that I must assume that Apple has something up its sleeve, like custom add in cards for Ray tracing acc or specialised multichip gpus. Or they could just make drivers for AMD GPUS again so that we who bought eGPUS can use them ..... I guess we'll all know in a few months.
 

singhs.apps

macrumors 6502a
Oct 27, 2016
659
396
Quick question, because I genuinely don't know : Who/What type of workflow uses more than 256GB of RAM?

Second question : When people go for Mac Pros, what percentage of users buy configs over 256GB ?

Everything points towards 384GB. Yes, it's a huge downgrade in terms of capacity. But I think Apple will let it go because above that is an extremely tiny market share.
I have 128 GB system ram and it’s not difficult to hit the limits for mid level to large scenes.
 
  • Like
Reactions: throAU

1096bimu

macrumors 6502
Nov 7, 2017
459
571
Now more than ever, we no longer need a desktop computer to have multiple graphics cards or terabytes of memory because of virtualization and cloud based compute resources. If you truly need absurd computation you can just run it remotely, you don't need 4 graphics cards in your desktop.
 

Wokis

macrumors 6502a
Jul 3, 2012
931
1,276
There's this persistent rumor that an Intel Mac pro will remain in the lineup precisely because Apple's approach to silicon can't, at least not yet, accommodate the requests in this space. I wouldn't rule it out.

Regarding general PCIe expansion, I'm expecting that to come even in an AS Mac Pro. You can already now achieve this with thunderbolt to PCIe enclosures. But unless something fundamental changes they won't support graphics cards.
 

Ethosik

Contributor
Oct 21, 2009
8,099
7,053
I am actually expecting a few CPUs here. I don’t think it will be M* family. I hope we see W* family with more functionality that M* series offers. There is a lot of different CPUs between a $6,000 2019 Mac Pro and a $50,000 version. I wouldn’t think we would just see one Extreme chip. I would hope they learned their lesson with the trash can Mac Pro. Apple pretty much apologized for the Mac Pro as much as their legal and PR teams would allow them to.
 

thenewperson

macrumors 6502a
Mar 27, 2011
985
896
I am actually expecting a few CPUs here. I don’t think it will be M* family. I hope we see W* family with more functionality that M* series offers. There is a lot of different CPUs between a $6,000 2019 Mac Pro and a $50,000 version. I wouldn’t think we would just see one Extreme chip. I would hope they learned their lesson with the trash can Mac Pro. Apple pretty much apologized for the Mac Pro as much as their legal and PR teams would allow them to.
Probably will just be binned versions up to the full set of cores. Can’t see them creating a family of chips for the MP.
 

Appletoni

Suspended
Mar 26, 2021
443
177
Pure curiosity before Mac Pro shows up later this year. Let's say a new chip's name for Mac Pro is M2 Extreme.

1. M2 Extreme will be a powerful chip but not in GPU because many workstations already can use more than 1 graphic card or can be used up to 4 or more. Mac Pro 2019 supports up to 4 graphic cards. Which means a new Mac Pro needs up to 4x M2 Extreme chips. If you are fine with just one M2 Extreme, then that's fine but how about those people who need more than just one graphic card? Are you gonna say you dont need more than 1 graphic card lol. There are reasons why many workstations support multiple GPU and you never know if Mac Pro becomes a beast in 3D market.
Adding an external GPU might be the option but Apple Silicon take advantages from unified memory and SoC so I doubt it.

2. The maximum RAM size is in question. With only one M2 Extreme, it might have up to 256gb or maybe up to 384gb based on M2's memory. Well, that's way lower than Mac Pro's maximum RAM size which is 1.5TB. Others can support up to 4TB. Even if they can use 4x M2 Extreme, it's barely 1.5TB. Yes, that's only IF they use 4 chips. And Yes, they might be able to add a lot of memory chips along with one M2 Extreme but can they really add those memories around the chip?

3. Will Mac Pro supports PCIe slots? What about upgradability and expandability? Currently, Apple is very hostile toward both area even for Mac desktops. If Apple wants Mac Pro to be another Mac Studio, what's the point? So far, I'm quite skeptical that Apple is willing to support PCIe slots and other upgradable parts.

I have no idea how AS Mac Pro will be look like but it shouldn't be another Mac Studio or Mac Pro 2013. If not, then what's the point of having Mac Pro? Mac Pro 2013 was a huge failure anyway.

We'll see after Apple announces a whole new Mac Pro...
Just put that thing inside a MacBook Pro 20-inch and it would be fine for chess.
 

MayaUser

macrumors 68040
Nov 22, 2021
3,098
7,070
there is a chance to update the Mac pro with Apple own SoC down the road? not to replace the whole thing every-time a new architecture is coming?
We will get 5nm M2 extreme or something like that...but in 2 years can we swap that 5nm SoC with a new 3nm M3 extreme SoC?! or we have to buy a new mac pro ?! im guessing the latter since the Ram is also in it, the gpu ...so nothing else remains
 

Ethosik

Contributor
Oct 21, 2009
8,099
7,053
Probably will just be binned versions up to the full set of cores. Can’t see them creating a family of chips for the MP.
We need more workstation functionality in the Mac Pro chips. Otherwise it will be a death to the pro market. It will be the trash can Mac Pro all over and I think that will start a decline in macs. M* is fine for consumers and some pro work. But not to the level of the Mac Pro. If it was simply an M* Extreme we would have seen it by now. I really hope it’s not a M* family. Need more PCIe lanes, expandable, larger RAM access and more than the M* support.
 
  • Like
Reactions: maxoakland and EugW

sunny5

macrumors 68000
Original poster
Jun 11, 2021
1,740
1,608
Now more than ever, we no longer need a desktop computer to have multiple graphics cards or terabytes of memory because of virtualization and cloud based compute resources. If you truly need absurd computation you can just run it remotely, you don't need 4 graphics cards in your desktop.
Both virtualization and cloud still not able to replace the entire workstation market. Beside, the internet connection is a huge bottleneck.
 

1096bimu

macrumors 6502
Nov 7, 2017
459
571
Both virtualization and cloud still not able to replace the entire workstation market. Beside, the internet connection is a huge bottleneck.
they don't need to replace the entire market, they just have to replace enough such that the rest is no longer profitable.

if all the resources are already on the server, even 1GBE is plenty enough for Remote Desktop.
 

theorist9

macrumors 68040
May 28, 2015
3,825
3,020
Upcoming competition from Nvidia will have GPUs in the 100 TFLOPS range and AMD in 75 TFLOPS range. Even without counting the specialised ray tracing cores. (and these theoretical figures actually maps almost perfectly to 3d rendering performance in SW like blender, redshift and octane). So this fall a freelancer can buy a AMD 7950 and have about 200 TF of rendering power. That is a magnitude better than what the best Mac can provide even assuming we had optimised SW for the Mac as well.
Pure speculation, but maybe they'll double up on the number of GPU cores in the upcoming M2 AS Mac Pro (if there is an M2 AS Mac Pro). That would give ~120 TFLOPS. Or maybe they've decided heavy-duty rendering isn't their market. Or maybe it's with the M3 that they'll really pull out the GPU stops (including hardware RT).

TFLOPS, SINGLE-PRECISION (FP 32)
M1: 2.6
M2: 3.6
M1 ULTRA: 21
M2 ULTRA: 29 (?) (EXTRAPOLATING FROM M2/M1 x M1 ULTRA)
M2 EXTREME: 58 (?) (EXTRAPOLATING FROM 2 x M2 ULTRA
M2 EXTREME WITH DOUBLED GPU CORES: 116 (?)
 

sunny5

macrumors 68000
Original poster
Jun 11, 2021
1,740
1,608
they don't need to replace the entire market, they just have to replace enough such that the rest is no longer profitable.

if all the resources are already on the server, even 1GBE is plenty enough for Remote Desktop.
It's actually slower than having a workstation because of the internet speed unless you have a rendering farm right next to you. First of all, if you put all resources on the server, then it will be slow to work with. Interal SSD's speed is already faster than internet speed. That's a huge bottleneck.

If not, then how come a lot of people still relying on buying workstations instead of using virtual or cloud services? Oh, do they even support Apple Silicon Mac and Final Cut Pro X?
 
Last edited:
  • Like
Reactions: maxoakland

Xiao_Xi

macrumors 68000
Oct 27, 2021
1,617
1,085
AWS is investing in enticing studios to move to the cloud.

Working from home seem to have changed companies' workflows.
 
  • Like
Reactions: theorist9

deconstruct60

macrumors G5
Mar 10, 2009
12,476
4,033
Both virtualization and cloud still not able to replace the entire workstation market.


They don't have to replace the entire workstation market. Just a significant fraction of the relatively very small subset that is running 4 GPU cards in a workstation. What gets replaced is a subset of the workstation market. The fallacy is that all of these market user bases are the same people fixed over a very long period of time. As technology changes users move from one segment to another. Several examples:

a. 90's the PC market was about 70+ % desktops and around 30% laptops. In 2022, that has flip flopped.

b. 70's more folks on Mainframes and Mini computers than personal computers. In 2022 , that has flip flopped.

c. Back when all Apple had in a Mac Mini or iMac was a laptop processor and the Mac Pro had the only 'desktop' process there was a much larger fraction of folks who "had to have" a Mac Pro. In 2022, a MBP 16" with Max SoC 'smokes' vast majority of standard configuration Mac Pro's sold back in 2010. And a substantial number of users have moved over. (e.g., lots of developers just use a high core count MBP to do work. ).


The largely fallacy is that the new thing has to be a 'killer' product. That it completely wipes out the old product. That really isn't true to do significant market damage to the old product. Just trimming off 20-30% can do serious damage to product viability in more than a few cases.

Workstation market ....

" ...

Our Best-Selling Workstation​

....
Up to 2x NVIDIA RTX™ A6000 or 2x AMD Radeon™ Pro W6800 GPUs2
...

So it can't do > 2 GPU cards. So missing out on the 3-4 GPU card market isn't going to knock Apple out of the workstation market because most workstations sold can't do that anyway.

The flaw here is that the workstation market is the Mac Pro 2019 configuration that is > 16 cores and starts at a W6800 duo. It is not.




Beside, the internet connection is a huge bottleneck.


The "internet" can go 15 or 15,000 feet. There is no inherent reason why 3 high end GPU cards couldn't go down the hall for another machine for communal batch/interaactive job use on a modern LAN set up for a variety of workloads.

4K video --> AV1 hardware compressor --> 10GbE ---> AV1 decompressor --> screen is not really a 'huge bottleneck'. At 24-48Hz screen refresh rates even more so. If sitting in some random basement with a '50Mb/s' home internet connection perhaps , but modern LAN speeds. Make a very short animation movie and play it back remotely on a screen. Netlflix, Amazon , etc serve up 4K to much slower end point network locations everyday.

Are there people 'stuck' on the far lower speed, high latency end points of the Internet? Yes. Is that everybody? No.

In the 90's - early 2000's there was a stampeded into more numerous , smaller servers. Average system utilization rates were low ( system running at 10-50% levels most of the time). Better virtualization and multitenant hosting infrastructure came in and there gradually was a wave of folks aggregating this stuff back up onto larger systems that can run higher average utilization rates ( 60-90%). In many cases it is more cost effective. If have 10 workstations with 4GPU that only run at max 4 throughput 20% of the time then have more than a few , largely idle expensive GPUs.

Over time , technology is going to push the one , big GPU towards being able to do more with fewer GPU packages. Same reason why the Mac Pro dual CPU package configuration dropped out after 2012. It isn't coming back going forward.
 

deconstruct60

macrumors G5
Mar 10, 2009
12,476
4,033
Pure curiosity before Mac Pro shows up later this year. Let's say a new chip's name for Mac Pro is M2 Extreme.

1. M2 Extreme will be a powerful chip but not in GPU because many workstations already can use more than 1 graphic card or can be used up to 4 or more. Mac Pro 2019 supports up to 4 graphic cards. Which means a new Mac Pro needs up to 4x M2 Extreme chips.

This is the unsubstantiated premise that sends the rest off into the road. The Mac Pro 2007-2012 Models had two CPU packages. Going forward past 2012 they only had one. All computatoinal workloads don't scale across GPUs. Many in the common Mac Pro user workloads space do , but some apps do not. For the ones that need one , 'big horsepower' GPU what Apple is doing is not disconnected from the technological trend line.

So there is no huge requirement that they have more than one SoC in the system. So no 2x , 3x , 4x SoC is not mandated by anything. Apple won't be able to sell every workstation configuration to everybody ... but Apple is not in the everything for everbody business ( e.g., dropped out of selling dual CPU package workstations around a decade ago. and are doing just fine (have profitable enough product to make/ship) in the workstation market.


The M1 Mini backslid on some configuration parameters ( total supported video screens out , max RAM capacity , etc. ) The iMac backlisde on parameters ( max RAM , in most cases supported video screens out , not even a direct 27" product (replaced by headless Studio). ) . None of that hurt Apple mac deskstop sales in any significant fashion. Sure there are some 27" iMac fans 'protesting' by not buying anything , but Apple as replaced them over the intermediate term with folks who do want to buy what is in the line up. The same people don't have to be constant.... Apple just needs the unit sales to grow and make more money. Stuff after that is 'gravy'.


There are also some major limitations here. First, without very deep and substantive changes to the kernel, macOS can't do more than 64 threads. So having more than 50 or so CPU cores really is a huge mismatch to where the kernel is going (and hence where macOS is going). So 4 * 40 CPU cores is way , way , way off in the swamp. Extremely likely not going to happen. The current MP tops out at 28 cores. So if they got to 40 , more powerful cores that would be a large net increases in CPU 'horsepower'. Trying to spin that as a 'fail' is a huge misdirection.

Second, some folks buy a Mac Pro with a 16-24 core and 580X/5550X and it totally fits their workloads. Not every Mac Pro workload has to be GPU bound. There is an relatively opposite problem/issue here is that some folks are being dragged into paying too much money for GPU cores they may not need. Pricing wise that could become a critical issue also once move far enough away from the entry price ( which is likely another demand lowering hurdle that moving up the MP entry price 100% was/is from 2008-2013 era. )





If you are fine with just one M2 Extreme, then that's fine but how about those people who need more than just one graphic card? Are you gonna say you dont need more than 1 graphic card lol. There are reasons why many workstations support multiple GPU and you never know if Mac Pro becomes a beast in 3D market.
Adding an external GPU might be the option but Apple Silicon take advantages from unified memory and SoC so I doubt it.

When Apple introduced the Ultra SoC and Mac Studio they demonstrated that the Ultra configuration often did better than a MP 2019 configuration that had 16 cores and a W5700. If Apple just doubled that then they'd be doing better than two W5700. So they would cover people who needed two W5700.

They also said that their most commonly sold configuration was a 16 core / W5700 . If there is a simple bell curve distribution on configurations that means half of the Mac Pros sold are at the same or under that 16 core / W5700 configuration. And that the 28 core / 2x W6800 duo is way , way , way out of on the fringe edge of Mac Pro's sold.


So again. Is Apple going to cover everybody? No. Are they going to cover enough to ship a viable product? Probably yes ( presuming the entry price is about the same). The current MP 2019 16c / 48GB RAM / W5700 goes for $8,699 . If the new entry Mac Pro is a M2 Ultra / 64GB that is $5,999 then that is more desired configuration at about $2,700 less. If Apple cuts the price that much they will probably sell lots more. That has a very good chance of offsetting the unit sales of the fringe 28 core / 2x W6800 sales. That will sell more Mac Pros, but might not bring in as much profit. But that will be OK because they'd also have "Extremes" to sell (which are not going to come cheap). Similarly, the 2 x W6800 Duo is $9,499 and add a Afterburner card it totals > $11K .. Similar move where Apple can chop the price $5K and they'll be able to get some people to move who have heavy, multiple concurrent stream 8K ProRes RAW workloads.


2. The maximum RAM size is in question. With only one M2 Extreme, it might have up to 256gb or maybe up to 384gb based on M2's memory. Well, that's way lower than Mac Pro's maximum RAM size which is 1.5TB. Others can support up to 4TB. Even if they can use 4x M2 Extreme, it's barely 1.5TB. Yes, that's only IF they use 4 chips. And Yes, they might be able to add a lot of memory chips along with one M2 Extreme but can they really add those memories around the chip?

Yes, but how many folks are out there. The bigger mismatch there though is not capacity , but lack of ECC. Going into the triple digits of GB RAM without ECC is going to get the system tossed from some workloads also (where data integrity is a top priority).

Again though, where is Apple's average/median MP customer configuration at? If the average is 128GB and they cover 256GB is that really going to be a huge problematical issue for product viability? It is a "bragging rights" problem , but bragging rights doesn't pay R&D costs. The real relevant issue is whether they can sell enough M2 Extreme SoC to make a profit or not. If Apple leaves 15K users on the sidelines, swaps in 5K more new users, and on aggregate makes 5% higher profit , then that is a slam dunk.

If Apple can avoid high capacities then they can avoid having to do ECC memory controller work. Pretty likely Apple would have to either roll ECC out of a couple more members of the Mac product line up to get a much bigger user base to amortize the costs over or they will want to skip it and allocate R&D resources to other parts of the SoC.

Apple has said the Mac Pro is in the single digits of Mac product market. Some folks presume that is 5% or higher. Pretty likely it is far closer to 1% . So if have a feature that 10% of 1% are using then basically rounds to 0%.

The 1.5TB stuff for the MP 2019 had an Intel tax on it. There were 24/28 core optons that were $2-3K less than Apple skipped to drop the more expensive versions into the Mac Pro. Apple slaps another Apple tax on top of the Intel tax. So Apple's margins were substantially higher selling the 1.5TB Xeon W 6200's . It is not clear at all if customer need was the dominate factor there at all. With M-series SoCs Apple makes good margins on the RAM costs, so they don't really need to goose the Max RAM capacity with a tax to make healthier margins. They are already are quite high.

Will they loose some customers at 256GB ? Yes. Do they need to completely solve that in the M2 generation? No. Even more so if willing to sell 4 year old MP 2019 models into 2023 and stop around 2024.



3. Will Mac Pro supports PCIe slots? What about upgradability and expandability? Currently, Apple is very hostile toward both area even for Mac desktops. If Apple wants Mac Pro to be another Mac Studio, what's the point? So far, I'm quite skeptical that Apple is willing to support PCIe slots and other upgradable parts.

This is basically vacuous . There are already over 50 cards that work with Apple M-series . "hostile" would mean that number would be a lot smaller.

Apple is hostile to three category of cards:

1. Overly depended upon early UEFI for proper initialization. There is no UEFI anymore in the new systems. So if the world view revolves around UEFI only then that is a bit 'hostile'. Apple isn't big on including others in the early boot process.


2. Storage cards that don't present as standard NVMe or SATA drives. If there is something funky the card wants to do at boot to present as a coherent drive then they are out. ( historically Apple put in hooks in legacy systems for things like SoftRAID to get included in standard Apple bootloaders. ). Once macOS is running if want to do an initialization there on macOS then that is fine. But that could be view as "hostile".

3. 3rd part GPU drivers. (pragmatically at this point dGPU drivers). There is nothing there.


[ macOS even on Intel hasn't been all that friendly to RDMA , very high end Networking cards. That isn't really 'new'. ]

if Apple was extremely hostile to PCI-e slots then they would be hostile to Thunderbolt. Provisioning those is a substantive feature of Thunderbolt. Apple has been a major proponent of it. It is the primary reason why there are >50 card ready already before a M-series Mac Pro even shows up. That has laid lots of ground work.



The bigger issue with the M-series so far is that Apple has been very keen to reduce the number of PCI-e lanes out of the SoC. They have taken the viewpoint that with PCI-e v4 that x4 lanes can do the same amount of bandwidth as x8 v3 lanes... so less lanes is OK.

That is the bigger issue. So instead of getting x64 top end lanes out, the number is going to dwindle back to x32 or x28 lanes that will be coupled to a PLEX PCI-e switch to be multiplied out. The laptop Mx Series SoC seem have x4 PCI-e v4 amount of lane provisioning. ( the Ultra gets a side-effect allocation of x8 and an over allocation of a TB controller ).

Apple does need more of a desktop focused building block. Doing "everything" with laptop opitmized large dies is going to be a mismatch.


I have no idea how AS Mac Pro will be look like but it shouldn't be another Mac Studio or Mac Pro 2013. If not, then what's the point of having Mac Pro? Mac Pro 2013 was a huge failure anyway.

The Mac Pro 2013 was being bought by the pallet right up until the last day it was sold. ( It wasn't the classic Mac Pro customers. It was places like MacStadium/Colocation customers that used it as a "Mac via a cloud service" system. It is/was a good fit for that in terms of compute density. ) .

The huge next Mac Pro issue likely does not hinge on the SoC as much as whether Apple wants to make it a literal desktop system ( has to fit on a desktop as opposed to beneath or beside it). Apple seems to have a "must fit in a 7" x 7" square" requirement for literal desktops. Mini, taller Mini (Studio) , foot of the iMac/XDR/Studio display stand . The Mac Pro saddled with that largely is what gets you the MP 2013.

Some rumors have been that Apple is doing something "half sized". At half the volume , there is a decent chance that literal desktop restriction is not in play. It will have lower shipping costs, but you'll still be able to add those $100/wheel option. :)

There was also a rumor of a one-slot-wonder that was an M1 SoC that probably is likley not to see light of day. Whether that is a prototype or Apple doing the barest minimum slot allocation, it is still indicative they are not trying to diretly revivie the MP 2013. That doesn't make any rational sense. The evolution for the MP 2013 is largely

MP 2013 --> iMac Pro ---> Mac Studio.

They already have something that largely matches that role. Why would they want another direct overlap with Mac Studio? Pretty likely though that the M2 "Utlra class" SoC for the Studio could/would overlap with the entry Mac Pro . Apple likely has to get the SoC run rates higher to fork off of just using MBP 14/16" SoC in a mash-up as the basic building block.


We'll see after Apple announces a whole new Mac Pro...


I suspect one of the other flaws here is that the expectation that the M2 Utlra is going to look almost exactly like the M1 Ultra. Using a the "Max"-class die as a chiplet is somewhat goofy. It makes sense used as a single die in a laptop, but once start moving it into the multiple chip packages it doesn't really scale. There is zero reason why M2 , M3 , M4 has to rigidly stick to the same size dies used in the same size patterns once new packaging technology and fab processes arrive.

There are two paths that would give a wider balance of Mac Pro (and Mac Studio) configurations.

1. Go to two more desktop oriented building block largish dies.

20 CPU cores + NPU cores + Pro class GPU cluster + x16 PCI-e v4 provisioning ( or x8 PCI-e v4 and 1-2 block TB controllers) + 2+ UltraFusion connectors.

bigger than Max GPU ( e.g., 40 GPU) cluster + display controllers + NPU cores + PresRes de/encode units , TB controllers + 2+ UltraFusion connectors


So would be able to do a variety of quads.

20 CPU + 40 GPU + 40 GPU + 40 GPU ( customers that are GPU core workload limited )

20 CPU + 20 CPU + 20 CPU + 40 GPU ( customers that are CPU core workload limited )

20 CPU + 20 CPU + 40 GPU + 40 GPU ( customers that need balance mix of both )



Could probably sell some of the CPU weighted dies as an alternative "Max" in some MBP 14/16" configurations. Or another desktop ( if bring back iMac large screen ). Is Apple going to create 7900/4090 'killer' GPUs that way? Nope. Will it be more completive that using laptop optimized building blocks? Yes. (going to cover more future upper mid-range. And that future upper midrange is going to be be 'good enough' for a wider set of folks. )


2. Even more modular. CPU + GPU + I/O in their own chiplets and fuse them together with ultra fusion. Could make "Max" class SoC out of chiplets rather than being medium-large monolithic die. That would keep the "Max"-class SoC sales volume for amortiziation of R&D and give a wider set of options for multiple CPU/GPU die packages .
Also breaks out the I/O into a separate chiplet so that can have a 'laptop' almost not PCI-e lane ones and 'desktop' more PCI-e lane one if that makes more sense.

If Apple went TSMC N3 and saved some power and then 'lost' small amount of power efficiency by changing Max monolithic die to chiplets then it could be an even 'wash' trade-off. It wouldn't be a battery win for the MBP's but it wouldn't be a battery loss either.




The other part is largely just software. If they carve out a Metal compute only subclass to the PCI-e driver framework classes then can weave back in GPGPU compute GPUs. I doubt they'd be provisioned directly in the Mac Pro case. but a PCI-e exapnader box was used by some back in Mac Pro 2009-2012 days to do 3-4 GPUs. Worked before ... could work again if just have a free x16 PCI-e v4 slot. ( v5 would be better but I don't expect Apple to go there in M2 generation. ) .
 

deconstruct60

macrumors G5
Mar 10, 2009
12,476
4,033
There's this persistent rumor that an Intel Mac pro will remain in the lineup precisely because Apple's approach to silicon can't, at least not yet, accommodate the requests in this space. I wouldn't rule it out.

The MP 2019 stretched out into 2023 by selling it 4 years. Pretty good chance. The Mac Mini 2018 has been stretched to 2022. So doing the same on the MP side would not be a big leap.

The rumor that Apple was going to come with a refreshed Xeon W-6300 (Ice-Lake) system with refreshed GPU MPX modules. That is not so creditable at this point. I wouldn't count on that. Or x86-64 macOS support not tripping a vintage/obsolete countdown clock at the end of 2022. Apple seems likely to continue selling the MP 2019 model into 2023, but it seems reasonably likely that they are not going to even ship a M-series options until 2023 anyway. That all 2022 could bring is just a 'sneak peak' at a new Mac Pro ; not an actual explicit shipping timeline.

Even on the x86-64 Windows PC workstation side most major vendors have skipped Xeon W-6300. Not really any good reason this late in 2022 for Apple to pick it up. If that had shipped back in Summer 2021 then maybe. But now it will get seriously beat up at the low end by the Gen 13 (raptor lake) and Zen 4 (Ryzen 9 7000 ) . It is beat up at the high end by the Threadripper W5000 series.

On the GPU front there are zero 3rd party GPU drivers for macOS on Apple Silicon. So it is pretty clear that is not a high priority ( It has been 2 years and nothing. Not even an abstraction class in the driver framework as a placeholder.). The 7900 is coming but also seems to be coming with a higher TDP. That makes a duo option unlikely (there never was a 6900 Duo anyway. And the W6900 was priced higher than the W6800 Duo; Apple had very low expectations of selling a significant number of those. Goosing crypto-craze driven profits margins yes. Unit volume, no. ). If there are not new Macs coming then new AMD drivers are likely not coming either. ( The 6800-6900 driver could have just falling out of a canceled W-6300 'plan b' system Apple ran up until there was working M1 Max/Ultra silicon that said they didn't necessarily need it. ). But even if Apple did AMD drivers for some narrow subset of the 7000 series, the cheapest option is to just stuff them into the existing (already paid for ) MP 2019 product. Really, users would not be getting a 'new' Mac Pro.


Regarding general PCIe expansion, I'm expecting that to come even in an AS Mac Pro. You can already now achieve this with thunderbolt to PCIe enclosures. But unless something fundamental changes they won't support graphics cards.

Even if Apple just used the laptop optimized chips with just x4 PCI-e v4 lane provisioning off the die. An Ultra SoC could easily have in the M2 generation a x4 PC-e slot ( or a cheesy one (or two ) x8 v3 slot if ran through an external switch). An quad-max would have three of those 'extra' units.

Similar chessy hack with the 'extra' TB controllers with x4 v3 provisioning behind them.

The major issue is someone demanding that the industrial design doesn't make the case "too small" to squeeze out the slot. The SoC have the build in capability to do something ( even if it is minimal ) just because there are multiple dies. ( so have extra provisioning because the laptop dies need some when used in single die package. WiFi , Ethernet , etc. ).
 

deconstruct60

macrumors G5
Mar 10, 2009
12,476
4,033
There is a "huge" crowd of creatives that need 3D rendering power. These guys are both freelancers and part of larger studios. They (we) need responsiveness during the whole pipeline from concept through modelling, lookdev, lighting etc. Freelancers usually buy a bang for the buck machine and for a few years that has been a 16 core AMD (1950, 2950, 2950, 5950) coupled with dual GPUs, atm that would be dual 3060 tis -> 3090 tis. This is a config that solves all the needs for the category of user.

There really isn't a huge gap here. Apple's Ultra in a Mac Studio is covering 16 core Mac Pro . On the CPU front , a M2 Ultra-class SoC would cover that historical range for CPU workload. Would it be a Ryzen 9 7000 'killer' ? No. But it would be competitive on a subset of that performance range.

Dual 3060's for a "double Utlra" isn't that far out of reach with some improvents on general GPU workloads. For super optimized , extremely Nvidia specific hardware RayTracing the MP 2019 isn't an great option either. Doesn't particularly make sense that the Apple Silicon system would have to go chase a niche that Apple isn't even in now before the transition. The MP 2019 ( and previous) have been a profitable enough product without that niche over the last two years. The new Mac Pro will not necessarily implode if it also does not have all of it. Chasing Nvidia super fans has not been an Apple priority for more than several years.


I hoped for a long time that we could get something "close enough" on the Mac side and it was pretty close with a iMac 27" and dual eGPUS with 5700 ->6900xt cards in. Sadly that is not a thing anymore and we are left with the stinking mess that is SoC. The studio ultra has a nice CPU but the GPU is a joke. For 3D rendering it is at best Nvidia 3060 levels...

Apple came out with new Apple GPU optimization educations and developer tracing/benchmarking tools at WWDC 2022. Several months after the Utlra's release. The notion that the Ultra GPUs have been completely optimized for is likely flawed. The tools shipped after the product. The first 6-9 months of Nvidia's brand new RT hardware was super optimized in all apps 2-3 months after it was released? Nope. AMD new release architecutres are all optimized across all software titles 2-3 months after it is released ? Nope. Not sure why expectations are being thrown at Ultra ( two sets of GPUs cores lashed over an extremely high speed bus ) where apps are suppose to come out the gate completely optimized in a short amount of time.

One of the dual edged swords of making app developers use extra low level APIs like Metal ( Vulkan , Direct 12 X low level) is that the optimization responsibility is shifted into the application developer hands. If they do a very good job then can do better than old style, more abstract OpenGL approaches. The down side is that they have to do all the work. If add optimizations at the mid-live insider of OpenGL then all the apps using that higher level API get the optimize ( fix once , impacts many apps). With Metal it is a much wider change propagation upgrade cycle. That isn't going to lead to faster spreading of optimized code across a broad set of apps,.


So, there desperately needs to come something that is not just a doubling of GPU perf. Upcoming competition from Nvidia will have GPUs in the 100 TFLOPS range and AMD in 75 TFLOPS range.

At what trade-off cost? They are also bringing TDP levels that makes doing multiple GPU cards harder to do for average currently installed/deployed cases. Some of what seeing here is common dual card setups being folded back into single card ( 'bigger' hotter than old single ) set ups. That is not entirely disconnected from what Apple is doing in moving 'old levels' of performance into lower aggregate system level power consumption.

Probably going to see a larger number of mid-range workstations that can now only take one quad (or more) wide monster module that does the work of two GPUs.

The top end AMP 7000 series has chiplet memory controllers and GPU cores pushed to TSMC N5 densities. The midrange product. Not so much.


Even without counting the specialised ray tracing cores. (and these theoretical figures actually maps almost perfectly to 3d rendering performance in SW like blender, redshift and octane). So this fall a freelancer can buy a AMD 7950 and have about 200 TF of rendering power. That is a magnitude better than what the best Mac can provide even assuming we had optimised SW for the Mac as well.

Longer term Apple really should at least loop back int GPGPU compute accelerators into the driver umbrella. The GUI stack is messier if they want to keep seamless ability to run native iPadOS/iOS apps on macOS. Intel has completely fumbled the ball by trying to build too diverse of a dGPU line up too fast from their iGPU foundation. Without the mature driver support there is almost not good way to do that. Time to go back and re-read the "Mythical Man month"; can't just throw money and bodies at the problem and get to a well constructed new system.

It is far more critical over the intermediate term that Apple has developers work out their mutichip Apple GPU issues first before moving further toward the top of the GPU performance range. Once have a solid foundation there, then move along.

Apple trying to go for a super top of the niche of a niche GPU card head strike, 'kill move' is goofy move. The Mac Pro as a product category has a load of economic viability issues and that isn't anywhere near the top 5 issues.
Gradually 'eat' the dGPU market from below worked well for Apple's M1 , M1 Pro , Max . Just evolve that over time.
The Axx-X (iPad Pro) SoC took several years to get to transition point in 2020. 2-3 years after 2022 isn't a show stopper for another jump.

[ Pretty likely going to see Intel 'retreat' back to covering laptop dGPUs ( which are always connected ReBAR/"higher shared memory" fashion and largely drop out out mainstream desktop ( gamer) add in card space for now.) ]


M3 , M4 generation this would be more of a deep seated issue that may rebalance more. Either Apple has better building blocks to build a even more extremely highly GPU core skewed SoC or they weave in more specialized GPU accelerators.

I mean, who would buy a machine that is at best 10% a fast as the competition and cost twice as much? It is so dumb a proposition that I must assume that Apple has something up its sleeve, like custom add in cards for Ray tracing acc or specialised multichip gpus. Or they could just make drivers for AMD GPUS again so that we who bought eGPUS can use them ..... I guess we'll all know in a few months.

If Apple lets GPGPUs back in the door it likely would be after WWDC 2023. It probably won't be a few months. I'd guess it would be 2024 before there was something that was solidly production qualified.

AMD GPUs as primary GUI driving GPUs. That likely isn't coming. Going forward Apple is kicking everyone non-Apple out of the kernel. There are no abstractions for 3rd party GPUs in DriverKit. Kernel extensions ( the current GPU drivers are deprecated even on the Intel side ). Trying to get all the App developers to do the proper, necessary optimizations for Apple GPU is like herding cats. ( Metal allow to somewhat randomly roam so they are collectively off into different directions). Apple allowing only Apple GPU drivers is a mechanism where they can focus attention and efforts after giving away some of the control. Without application evolution this isn't going to work. ( flogging old x86 OpenGL OpenCL architected code to go faster isn't going to work in the long term. Apple needs new freshly architected apps (and drivers ) that they don't directly control. )

Since Apple is so keen on AR/VR products as a growth category there is probably energy efficient RT coming out of that stack that could/would likely be folded over into the Mac (and iOS/iPadOS) SoC operating systems. Its is doubtful the Mac Pro is the primary driving product there. Any hardware RT system that didn't met the headset power requirements would be relatively useless for Apple as a whole. The evolution driver product there is likely outside the Mac space. Also Apple's Tile cache soak up a decent amount of die space. Apple is likely waiting for some N3 (or better) fab shrink to open up transistor budget for that.


Apple actually could bring MP 2019 16 core , W5700 performance prices down (not up). [ post #20 above] . Pretty good chance Apple is likely going to trade-off selling more mid-range units for unit losses at the very tip-top of the old Mac Pro configuration line up.

For the M2 generation I suspect best chance is some incremental progress where in some configurations trade off dumping CPU cores for adding more GPUs with a permutation of chiplets. For workloads that are more GPGPU embarrassingly parallel skewed workload, an 8P+4E CPU is likely good enough as an application code foundation.
 

Boil

macrumors 68040
Oct 23, 2018
3,413
3,086
Stargate Command
@deconstruct60

Thoughts towards the whole N3 -> N3E -> N3P -> N3S -> N3X process progression...?

Might we see Apple introducing a new Mn-series SoC, and updating that SoC series as the node improvements become available...?

So rather than a yearly / 18 month product cycle, we see a three year product cycle for a particular Mn-series...?

Launch the 13" laptops with the Nn node...

Launch the 24" iMac & Mac mini with the NnE node...

Launch the 14" & 16" laptops, a possible 27" iMac, & the high-end Mac mini (Mn Pro SoC) with the NnP node...

Launch the Mac Studios with the NnS node...

And finally, launch the Mac Pro with the NnX node, which takes all the improvements of the four previous nodes and gives the end user an actual "desktop/workstation-class" SoC...?

As a particular Nn* node improves, existing products can be "refreshed" to use the newer node process (as makes sense, obviously the NnX node process would be reserved for the Mac Pro)...?

Maybe we see two variants of the Mn Max SoC; a laptop-oriented one for the 14" & 16" MacBook Pro laptops & a desktop-oriented one for the Mac Studio & Mac Pro headless desktops/workstations...?

I do like the chiplet idea though, rather than an end user being locked into more CPU or GPU cores than they need, Apple can offer a variety of configurations that let the end user get the cores they need without having to also have cores they do not...

Regarding PCIe slots...

Dropping discrete GPUs & not needing an Afterburner card, that right there cuts 48 lanes out of the picture (x16/x16 for the GPUs & x16 for the Afterburner card); the only thing really using up a bunch of PCIe lanes at that point would be a hardware RAID card with multiple M.2 SSDs...? Most everything else (A/V I/O & networking) would not have a need for x16 lanes...?

Hoping a preview/sneak-peek of the ASi Mac Pro will come along sooner than later and reveal a bit more of the overall plans for ASi Macs going forward...!?!
 

ILoveCalvinCool

macrumors 6502
Feb 21, 2012
258
595
Pure curiosity before Mac Pro shows up later this year. Let's say a new chip's name for Mac Pro is M2 Extreme.

1. M2 Extreme will be a powerful chip but not in GPU because many workstations already can use more than 1 graphic card or can be used up to 4 or more.


Wouldn't it be more about the number of cores than the number of cards? The M1 Ultra is just a number of smaller M1 chips stitched together.
 

theorist9

macrumors 68040
May 28, 2015
3,825
3,020
They don't have to replace the entire workstation market. Just a significant fraction of the relatively very small subset that is running 4 GPU cards in a workstation.....

Workstation market ....

" ...

Our Best-Selling Workstation​

....
Up to 2x NVIDIA RTX™ A6000 or 2x AMD Radeon™ Pro W6800 GPUs2
...

So it can't do > 2 GPU cards. So missing out on the 3-4 GPU card market isn't going to knock Apple out of the workstation market because most workstations sold can't do that anyway.
Sure. But you don't need to go to 3-4 GPU's to see a big gap. Let's consider just dual-GPU machines. When the AS MacPro comes out, it's going to be competing with dual 4090 (est. 160 TFOPS), dual 4090 Ti (est. dual 180 TFLOPS) or dual Quadro cards with whatever the sucessor is to the A6000 (comparable to the 4090/4090Ti). And unless Apple does something pretty fancy with their GPU's, those are going to leave the AS Mac Pro in the dust in GPGPU compute,

Indeed, an M2 Extreme (est. 60 TFLOPS, unless they do something fancy) isn't even going to be competitive with a single 4090 (est. 80 TFLOPS) in general GPU compute.

Now maybe it doesn't need to be. But let's be accurate in characterizing where the gap will actually be. It's won't be vs. 4 X GPU machines. That's a straw man. Rather (again, unless they do something fancy), it will fall short vs. high-end 1x GPU machines in GPGPU compute!

So the argument you'd actually need to make isn't that only a tiny fraction of the workstation market is running 4x GPUs. Rather, you'd need to argue that only a tiny fraction of the workstation market will run anything more powerful than a single 4080 Ti (est. 60 TFLOPS) (or the Quadro equivalent).
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.