New Mac Pro may not support PCI-E GPUs
The forthcoming Mac Pro is reportedly unlikely to add support for GPU PCI-E cards, as well as not allowing user-upgradeable RAM.
-xl-xl-xl.jpg)
The New Mac Pro could look like the old one.
Apple Silicon Macs have lacked support for external GPUs from the start. Now it appears that the M2 Mac Pro cannot change that.
According to Bloomberg's Mark Gurman, the Mac Pro "may" lack upgradeable GPUs, and this adds to its expected lack of modularity and expansion options.
"That will leave storage as the main user-upgradeable component in the new Mac Pro," continues Gurman, "which will have the same design as the current, Intel model."
"The big difference between a Mac Pro and a Mac Studio -- in addition to M1 Ultra to M2 Ultra," he says, "should be performance from more cooling."
It's been previously reported that the Mac Pro will not allow after-market RAM upgrades, because the memory is part of the Apple Silicon processor.
While the Mac Pro remains on Intel processors, there are circumstances where this most expensive Mac can be effectively equaled by the lowest-cost one, the Mac mini.
Read on AppleInsider
The New Mac Pro could look like the old one.
Apple Silicon Macs have lacked support for external GPUs from the start. Now it appears that the M2 Mac Pro cannot change that.
According to Bloomberg's Mark Gurman, the Mac Pro "may" lack upgradeable GPUs, and this adds to its expected lack of modularity and expansion options.
The next Mac Pro may lack user upgradeable GPUs in addition to non-upgradeable RAM. Right now Apple Silicon Macs don't support external GPUs and you have to use whatever configuration you buy on Apple's website. But the Mac Pro GPU will be powerful with up to 76 cores.
-- Mark Gurman (@markgurman)
"That will leave storage as the main user-upgradeable component in the new Mac Pro," continues Gurman, "which will have the same design as the current, Intel model."
"The big difference between a Mac Pro and a Mac Studio -- in addition to M1 Ultra to M2 Ultra," he says, "should be performance from more cooling."
It's been previously reported that the Mac Pro will not allow after-market RAM upgrades, because the memory is part of the Apple Silicon processor.
While the Mac Pro remains on Intel processors, there are circumstances where this most expensive Mac can be effectively equaled by the lowest-cost one, the Mac mini.
Read on AppleInsider
Comments
But I wouldn't be surprised if Apple really doesn't support GPUs in any form at all. They have made *such* a big deal out of the benefits of shared memory between the CPU and GPU. I'm inclined to think they really mean it. I really wonder if the next step for the M series will be distinct CPU and GPU chiplets connected by UltraFusion, thereby allowing pro users to better pick the right balance of CPU and GPU power for their needs. The vast majority of volume might still be fully integrated, single die SOCs, but for the 'pro' market splitting CPU and GPU onto different chiplets could be the way to go.
I know people might say they won't do it because of low volume, but I don't think that's such a big issue. These hypothetical 'pro' chiplets would still have the same CPU, GPU, and NPU core designs, so they benefit from economies of scale there. It's the 'glue' (UltraFusion) that's really unique, but Apple has already shown a willingness to do that in small volumes. Plus, 'pro' customers are accustomed to paying a ton for high-end chips -- those Xeons ain't cheap.
I think part of the GPU story is being missed, but we will see. Apple knows they have a hole to fill in preview rendering and they have been putting a lot of effort in to making metal ready on the software side. No idea what the plan is, but I wouldn’t be surprised if they create some type of rendering acceleration card that might not work as a normal GPU. I can’t imagine they will say they have the embedded GPU– end-of-story. As great as that GPU is, it isn’t quite powerful enough for people buying $50K machines which need a minimum 4X more GPU performance. I’ve thought that either they are getting AMD support for compute or maybe something out of the 2020 partnership with Imagination Technologies. Imagination has been helping with hardware ray tracing. Maybe they also have an ace up their sleeve. Originally Apple was building a discrete GPU for these machines code named Lifuka, but it was rumored to have been canceled for being too slow at the tasks it needed to perform. The Imagination contract was signed pretty close to the time that GPU was rumored to have been canceled. This is all conjecture, but I think we will see a surprise when the Mac Pro is launched.
It would, however, be very Apple-like to use an UltraFusion-like interconnect (which is similar to AMD's Infinity Fabric), so the MPX options at launch would be limited and expensive.
"76" is, I believe, the maximum GPU cores in an M1 Ultra chip, so while he didn't pull the number out of thin air, it doesn't actually lend any credence to his "report". But, like many of Gurman's "reports", it's couched in language that appears to say a lot but actually commits to nothing. I think he sometimes has information, and sometimes he doesn't and writes authoritatively regarding what he thinks is plausible; but he's also carefully ambiguous, and never distinguishes between the cases. And, since it's Bloomberg, there are no journalistic standards to adhere too, so ...
The man is a master of appearing to have far more knowledge about what's going on than he does, and I think a lot of his "inside information" comes straight from inside his head.
Apple hates modularity. Always has. Heaven forbid users want to buy expansion not made by Apple. They love to pressure you into buying more at the start “just in case” because you can’t add more later. This seems like it was always the plan. Nvidia is killing it lately and for years but Apple has to block them because they can’t compete.
I feel like Apple only went with Xeon Intel Mac Pros because they hated the product and wanted it to be as unappealing and overpriced as possible. It seems obvious that they always intended to kill off PCIe despite it being such a massive ecosystem that professionals rely on.
This company is really pushing its luck. The do great one day then completely screw you over the next. I love my iPhone 14 Pro Max—a really great product far ahead of the competition in quality (IMO) and price—but despite advances in some aspects their other products seem to be getting gradually less and less appealing to me. I’ll see what the M2 Max Mac Studio looks like, if there even is one, because that’s the only Mac that I’d possibly consider at this point.
I suppose I’m mostly concerned it will be another Apple so-so attempt where it will be kind of almost enough, but not truly competitive, forcing us once again into a situation I was hopeful Apple Silicon would resolve when it was announced.
Currently, from what I’ve seen, it doesn’t much matter if it is 72 GPU corers, 128 GPU cores, or 256 GPU cores, etc. it won’t be competitive (or even really close to the Intel Mac Pro) without some GPU piece we’re currently missing.
Kind of a wild and shaky rumour here, but unfortunately matches what I’ve been anticipating. My read on it is that if you need serious GPU power, a PC is in the cards for the next couple of years. Then maybe.
1) the same thing we have with the other Macs: storage, etc.
or, if Apple really wants to innovate:
2) a new motherboard framework that acts as a “fabric” which ties together multiple SOCs, with slots that you can add more SOCs after purchase. Each SOC would contain CPU, GPU, and RAM and that would be how you expand. Apple would of course charge crazy money, laugh all the way to the bank, and yet it would still enable customers to get what they want.
Gurman simply took the 38 cores of the M2 Max and doubled it as he thinks the Mac Pro will be using an M2 Ultra.
Apple has a very high bandwidth interconnect to pair two M1 Max chips. But who says that this interface can only pair two identical chips? Apple could make a companion chip that was simply made up entirely of GPU cores. So you could have an M2 Ultra with 12 CPU cores, the NPU and encoders/decoders from the M2 Max and have an M2 GPU with a whopping 128 (or more) cores.
Or if Apple went with a 4 die chip (the rumored Extreme version) then 2 x M2 Max along with 2 x M2 GPU and you’d have 24 CPU cores and upwards of 300 GPU cores.
I don’t know why people assume Apple can only connect two identical chips when they make an Ultra version.