Me too! The only thing that could really change my mind on that now is the release of a larger iMac that includes the Ultra chip.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
M3 Ultra Mac Studio Rumored to Launch in Mid-2024, But No Sign of New Mac Pro
- Thread starter MacRumors
- Start date
- Sort by reaction score
No, and no.
So much presumption in this article. Also showing that "rumors" just tend to be people replaying the past as a guess for what the future may bring.
Ultra is just a name, a branding for marketing. Have you noticed Intel is using it for the Meteor Lake line of chips?
And there is quite a bit of illogic in the idea that if a future M4 uses the N3E process, that then all N3E chips are "M4". I think if you look up a list of fallacies in a philosophy dictionary you'll find one that fits the statement by the original poster.
"Pro", "Max", and "Ultra" are just brands, an attempt by Apple to make words magical.
If TSMC's N3E process is used for such, the purpose of doing so will be because N3E has higher yield rates (if indeed that is true). Any small performance gain by using N3E instead of N3B is not important compared to the doubling of the processors that prior "Ultra" SOCs have brought.
Yes, but it's a good educated guess to assume that the M3 Ultra, too, will be two interconnected M3s Max.
Not true. It was rumored that was the case, but neither Apple nor TSMC said anything of the sort. Production on that node started in December of 2022. And yes, while the yields were initially low, they have improved and will continue to improve. And since basically no one else is using it, all production goes to Apple. So they’ve had plenty of time to grow their supply.
Apple does not hate N3B, in fact, they worked with TSMC during its development. It’ll be interesting to see which process Apple chooses to use next and when.
Agreed… Unless Apple decided to completely redesign the M3 Max for the N3E node, then yeah, they’re going to stick with the M3 Max using the N3B node for the M3 Ultra.
There are clear disappoints with the new Mac Pro and it isn’t only in the lack of a “quad’ processor. The former Mac Pro allowed people to configure a lot of processor, ram, graphics OR ANY COMBINATIONS therein. The move to Apple Silicon really makes this more challenging of course, and is not a surprise ultimately.
However, the real surprise for me was the price increase. It seems to me they should have lowered the Mac Pro by $1000 (from $6000) To $5000 not raised it to $7000. It seems reasonable to me that for $1000 you would get the Mac Studio + Mac Pro case, but the extra case/PCI slots are a $3000 premium now. That just seems almost as exorbitant as a $1000 stand.
However, the real surprise for me was the price increase. It seems to me they should have lowered the Mac Pro by $1000 (from $6000) To $5000 not raised it to $7000. It seems reasonable to me that for $1000 you would get the Mac Studio + Mac Pro case, but the extra case/PCI slots are a $3000 premium now. That just seems almost as exorbitant as a $1000 stand.
How do you know they’re blaming anyone? The reality is they understand the market is too small.
Recall that they also had a period where they struggled with their laptops...remember the 2016-18 MBP's with their butterfly keyboards and touch bar (OK, some liked the touch bar).
Having said that, it does make sense that the highest-end desktops are a struggle for them under AS, since remember that AS for Mac is built upon their experience creating AS for the iPhone and iPad, so the closer they get to high-end desktops, the futher they get from their original base expertise.
You saw that explicitly with their inability to bring a 4x Max (the so-called "Extreme" chip") to market. [At least that's the case according to the rumors--they say the reason we didn't see an Extreme chip in the MP wasn't because they weren't interested in creating such a beast, but rather b/c they tried but ran into too many difficulties.]
I expect they will get better at this as they gain more desktop experience with AS.
OK, don't get me wrong here. I understand that the AS MP was a disappointing product, in a few different ways.
And like you, I also didn't really see the point of it myself—until I read a post from an audio producer who said he was very happy that Apple offered it. He explained that he was using a 2019 Intel MP, and had about $30,000 worth of PCIe audio cards that would need to be transferred to the new machine. With the Mac Studio, he would have had to buy separate external boxes for them.* With the the MP, however, he was able to just slot them inside.
That's who the AS MP is for.
To determine how much sense it makes to offer an AS MP just for that market, you'd need to know what portion of Intel MP buyers use PCIe cards. I don't have that data, but Apple certainly does (or at least has a good estimate).
*Professional-grade PCIe 4.0 expansion boxes aren't cheap. Two of these would hold just as many PCIe cards as the MP (not including the MP's I/O card) (two full-length plus four half-length), and they're $1,200 each, for a total of $2,400:
Echo III Rackmount System (3 PCIe slots for full size cards + Module support)
Rackmount Thunderbolt 3 adapter box with three PCIe expansion slots for audio, video, fibre, RAID and ethernet PCIe cards.
www.sonnetstore.com
Last edited:
The Mac Pro is effectively dead, and once again it's Apple's fault. They had a chance to absolutely annihilate the competition & introduce an entirely new paradigm, instead they tack on some useless (heavily bandwidth-constrained) PCI slots onto a Mac Studio and charge an extra $3k. Absolutely insane. It was either a total stop-gap or a zero-effort send-off to accommodate the few thousand people who might want this thing.
I am hoping against hope that they have one last revolution left in them with this platform, although chances are almost zero. I am hoping the fact that they kept the same case (with the same beefy power supply) and went to all the trouble of developing the MPX modules means that something incredible is on the way.
What was needed to take this thing to the next level -
- Figure out the quad chip
- Leverage MPX modules to allow for expansion, maybe even additional M2 modules with extra RAM and all. Perhaps even special "MPX Compute modules" that have specialized silicon for different tasks.
- Figured out a way to allow for third party GPUs, as even what I imagine an M3 Ultra will end up being will struggle to compete with the best of what AMD and Nvidia have to offer right now, let alone 6 months from now.
- Worked with third parties (like last time) to help them build out an ecosystem of components that would integrate tightly with Apple Silicon and increase performance...maybe related to #2.
- Make it a bit easier to upgrade the on-board SSD storage
This would have been the hard thing to do, but it would have been a game-changer for many professionals and enthusiasts. They could have built a machine with potential to be insanely powerful, expandable and flexible. Instead, they did less than the bare minimum. I can only hope that's because they're either working on some/all of the above, or they're hoping people just forget about the whole thing.
I am hoping against hope that they have one last revolution left in them with this platform, although chances are almost zero. I am hoping the fact that they kept the same case (with the same beefy power supply) and went to all the trouble of developing the MPX modules means that something incredible is on the way.
What was needed to take this thing to the next level -
- Figure out the quad chip
- Leverage MPX modules to allow for expansion, maybe even additional M2 modules with extra RAM and all. Perhaps even special "MPX Compute modules" that have specialized silicon for different tasks.
- Figured out a way to allow for third party GPUs, as even what I imagine an M3 Ultra will end up being will struggle to compete with the best of what AMD and Nvidia have to offer right now, let alone 6 months from now.
- Worked with third parties (like last time) to help them build out an ecosystem of components that would integrate tightly with Apple Silicon and increase performance...maybe related to #2.
- Make it a bit easier to upgrade the on-board SSD storage
This would have been the hard thing to do, but it would have been a game-changer for many professionals and enthusiasts. They could have built a machine with potential to be insanely powerful, expandable and flexible. Instead, they did less than the bare minimum. I can only hope that's because they're either working on some/all of the above, or they're hoping people just forget about the whole thing.
Apple knows exactly what the high-end workstation users need. All they had to do was keep making the classic Mac Pro until the bauxite mines around the world were exhausted. They thought they knew better. Apple's hubris knows no bounds.
I saw that argument which another audio producer countered "the thing is a lot of people are moving to external boxes, and we have easy to work with enclosures now for our studios that the need for internal BlackMagic is not as big as it used to be"
OK, it sounds like some people have moved to external boxes and others haven't.
Given this, we can still say 'those producers who don't like external boxes for their PCIe cards are buying the MP rather than the Studio'. At least in my view, that's a more accurate characterization than "the only thing you can really do with it being adding storage and BlackMagic cards."
Last edited:
If they had a release schedule that mean't products were released when the entire range was ready and had been tested properly, Apple marketing would have a complete meltdown as there would be months or more likely a couple of years between product launches. Similar story for those who report on Apple in the media. Their entire business model is now built around about something new to talk about every few months, with rumours in between to keep the eyeballs on the page. The whole business is marketing driven, with the software and hardware engineers at Apple, desperately trying to keep up with the artificial cadence demanded by marketing and execs.
I understand but isn’t the $2400 for the external chassis still significantly less than the premium of a Mac Pro over a comparable Mac Studio?OK, don't get me wrong here. I understand that the AS MP was a disappointing product, in a few different ways.
And like you, I also didn't really see the point of it myself—until I read a post from an audio producer who said he was very happy that Apple offered it. He explained that he was using a 2019 Intel MP, and had about $30,000 worth of PCIe audio cards that would need to be transferred to the new machine. With the Mac Studio, he would have had to buy separate external boxes for them.* With the the MP, however, he was able to just slot them inside.
That's who the AS MP is for.
To determine how much sense it makes to offer an AS MP just for that market, you'd need to know what portion of Intel MP buyers use PCIe cards. I don't have that data, but Apple certainly does (or at least has a good estimate).
*Professional-grade PCIe 4.0 expansion boxes aren't cheap. Two of these would hold just as many PCIe cards as the MP (not including the MP's I/O card) (two full-length plus four half-length), and they're $1,200 each, for a total of $2,400:
Echo III Rackmount System (3 PCIe slots for full size cards + Module support)
Rackmount Thunderbolt 3 adapter box with three PCIe expansion slots for audio, video, fibre, RAID and ethernet PCIe cards.
Let's call it $2500 including a couple of TB4-certified cables. So it's $7,000 for a clean all-in-one solution vs. $6,500 for a messier 3-box solution. I can see someone for whom the $500 isn't a substantial amount of money going for the former.
Last edited:
I don't think that would work for Apple, for both practical and marketing reasons.
Practical: If they continued to have the MP be an Intel box, they would need to continue to support Intel on MacOS, which significantly complicates their future OS development work. I'm sure they're looking forward to going AS-only on MacOS.
Marketing: It's hard to market how superior AS is to Intel, if your halo product continues to be an Intel machine.
This. Linux in a powerful workstation is really hard to top. Apple could have kept up in this market - they have the resources, obviously - but they probably made the calculation that it's just not worth it. Not enough money in this compared to just pumping out iPhones and MacBooks.
The funny thing is that more "professionals" use Macs these days than back during the cheesegrater era. They're all just devs and creatives using MacBook Pros. This isn't the high end workstation market, but it's probably enough to give Apple that sort of prestige halo marketing ("buy this because pros use it") that the Mac Pro used to offer.
Two scenarios other than the Mac Pro being discontinued…
1) The m3 ultra may be reserved for the Mac Pro moving forward. The studio will get the M3 pro and M3 max moving forward.
2) The studio will get the M3 max, but Apple has the M3 Extreme under development for the Mac Pro and has done a pretty good job of keeping it under wrap.
Thoughts and opinions?
1) The m3 ultra may be reserved for the Mac Pro moving forward. The studio will get the M3 pro and M3 max moving forward.
2) The studio will get the M3 max, but Apple has the M3 Extreme under development for the Mac Pro and has done a pretty good job of keeping it under wrap.
Thoughts and opinions?
If 1 happens they'll be accused of downgrading the Studio. Of course Apple isn't above arbitrarily doing such things, but I think they wouldn't in this case because the Studio seems to be a more popular product.
Nope, not gonna work. Beside, M3 Extreme is already unrealistic since it's not just about connecting 2 or 4 chips. M2 Ultra is already too expensive and too difficult to manufacture due to its die size. That's the limitation of SoC and multi-die. Chiplet is the only option.