Here's Why Apple is Unlikely to Release an M4 Ultra Chip for Macs

[chucker23n1]

Once upon a time, kids…

Keep in mind that came at a time when there was a Macintosh Performa 5420CD, not to be confused with the Macintosh Performance 5410CD, based on the Power Macintosh 5400. Am I supposed to get an LC, a Performa, or a Power Macintosh? Who knows! The poor salesperson certainly doesn't. Plus, Apple at the time relabeled and sold CD-ROM drives, scanners, printers, and even digital cameras. Why? Hard to say.

Steve dramatically simplified the question of "what should I buy, and why".

In 2025, I don't think that question is very difficult to answer. For a Mac, the answer is probably a laptop. And probably the Air. If you need more oomph, you'll know you do, and in that case, or if you really care to get a better display, get the Pro. Then pick a screen size. Then pick the internals. Not that hard. If you really want a desktop, get the iMac if you want to be cute, or the mini. If it doesn't have enough oomph, get the Studio. If you have extremely specific requirements (you do not), get the Pro.

Sure, that's four desktops instead of two, but the iMac is more of an AIO than a desktop, the Studio is "mini, but more", and the Pro is very, very specialized. Nobody is confused "should I get the iMac, or the Mac Pro?". Someone might be confused "is the mini enough, or do I need the Studio?", which really mostly depends on how high-end your CPU/GPU/RAM requirements are.

 

[chucker23n1]

How long until they hit a point without any more meaningful progress, if they really release a new generation every one or two years? Will the M11 really be much faster than the M10? What else can they to except making chips bigger and denser?

"Who will ever need more than 640 KB RAM?"

 

[Veinticinco]

Keep in mind that came at a time when there was a Macintosh Performa 5420CD, not to be confused with the Macintosh Performance 5410CD, based on the Power Macintosh 5400. Am I supposed to get an LC, a Performa, or a Power Macintosh? Who knows! The poor salesperson certainly doesn't. Plus, Apple at the time relabeled and sold CD-ROM drives, scanners, printers, and even digital cameras. Why? Hard to say.

Steve dramatically simplified the question of "what should I buy, and why".

In 2025, I don't think that question is very difficult to answer. For a Mac, the answer is probably a laptop. And probably the Air. If you need more oomph, you'll know you do, and in that case, or if you really care to get a better display, get the Pro. Then pick a screen size. Then pick the internals. Not that hard. If you really want a desktop, get the iMac if you want to be cute, or the mini. If it doesn't have enough oomph, get the Studio. If you have extremely specific requirements (you do not), get the Pro.

Sure, that's four desktops instead of two, but the iMac is more of an AIO than a desktop, the Studio is "mini, but more", and the Pro is very, very specialized. Nobody is confused "should I get the iMac, or the Mac Pro?". Someone might be confused "is the mini enough, or do I need the Studio?", which really mostly depends on how high-end your CPU/GPU/RAM requirements are.

I was, of course, making a simplistic point.

You've explained it well, but then here, we're all much more clued up on the differentiation. For the average consumer, asking reasonable questions in an Apple Store from an equally confused and clueless retail employee?

Answer this though. Of the 2 Apple business models, 1998-2013 (taking us to the Pro Trash Can, allowing 2 years for Jobs' final work to pan out) vs. 1985-1997/the current state of fragmented affairs, which do you think is much more geared towards prolonged long-term success?

 

[chucker23n1]

I was, of course, making a simplistic point.

You've explained it well, but then here, we're all much more clued up on the differentiation. For the average consumer, asking reasonable questions in an Apple Store from an equally confused and clueless retail employee?

Answer this though. Of the 2 Apple business models, 1998-2013 (taking us to the Pro Trash Can, allowing 2 years for Jobs' final work to pan out) vs. 1985-1997/the current state of fragmented affairs, which do you think is much more geared towards prolonged long-term success?

I really don't think the current state is comparable to 1997. Back then, they had weird notions of "we have to rebadge them as 'Performa' for certain market segments" and "Japan needs slightly different models". (For example, there was a Color Classic II that… only shipped in Japan and Canada.) They overcomplicated things. Add to that that they do many of their sales direct to consumer, via their online store and physical retail stores.

I'll also point out that the simple matrix Jobs put up was only true for a brief period of time. In 2000, they introduced the Cube, leaving them with three desktops. In 2005, they introduced the mini. In 2008, the MacBook Air, with three laptop lines for a while.

As far as long-term success, the only Mac I'm currently unsure about is the Pro, but that isn't about customer confusion. Nobody is unsure "should I get the Pro instead?". They'll know when they need it. The Pro, however, has an uncertain future because so few people need it.

 

[MacLC]

Could you guys confirm whether Apple produces M3 Max with and without the interface for the UltraFusion connector?
I would expect that they bin the chips and only put the ones in the Ultra, where the connector part is okay.


And finally I wonder if they don't need a M4 Ultra as the M5 Ultra comes next year and it's not worth their time to make the M4 Ultra in-between.

That makes sense. It could be yield issues or it could be imitating Intel's old Tick-Tock scheme of alternating generations of die shrinks and optimizations.

Once upon a time, kids…

It was so refreshing when Steve Jobs canned all the confusing models such as the old Centris, Performa, PowerMac 6x00, 7x00, 8x00, 9x00, etc so people could simply buy one of 2 models in their category.
Back then they had consistent hardware updates. Those were the days.

 

[ZiBart]

Or just drop the Pro. The usual benefits of a tower workstation don't apply to the Pro - massive amounts of memory chips, multiple large GPUs.

Or that. I only said cancel the Studio because Apple retails the MPro for so much more where they could blend the added cost of the Ultra SOC. But yeah, keep the Studio and cancel the MPro would work even better. I guess i also thought that Apple could open up the MPro for add ons like the old days but these days its all baked in or can be done via TB4/5.

 

[Riot Nrrrd]

In 2025, I don't think that question is very difficult to answer. For a Mac, the answer is probably a laptop. And probably the Air. If you need more oomph, you'll know you do, and in that case, or if you really care to get a better display, get the Pro. Then pick a screen size. Then pick the internals. Not that hard. If you really want a desktop, get the iMac if you want to be cute, or the mini. If it doesn't have enough oomph, get the Studio. If you have extremely specific requirements (you do not), get the Pro.

It’s more nuanced than that (in certain cases). For example, I (stupidly) bought an LG 34” 5K2K Ultrawide monitor. I can’t buy any M4-based Mac of any kind, because Apple doesn’t support the default “ideal” resolution on Ultrawides - 3840x1620 @60Hz, scaled HiDPI.

Furthermore, if I bought an M4-based Mac mini in addition to the resolution issue I’d probably have problems with my USB-A devices, as I have more than 2 of them - so I’d need to use the rear ports on the mini and there’s a known random disconnect issue with that Mac.

The problems with the actual produced M4 Macs are incredibly frustrating, as the M4 seems like such a great chip line. Even without an M4 Ultra.

 

[Plerf]

Then why not toss the M3 Ultra in the Mac Pro and do that self handslap thing you do when you say something like “job done”?

 

[o9p0]

How long until they hit a point without any more meaningful progress, if they really release a new generation every one or two years? Will the M11 really be much faster than the M10? What else can they to except making chips bigger and denser?

*noticeable progress. the scaling and performance gains still probably look relatively the same when averaged out over time.

 

[chucker23n1]

Then why not toss the M3 Ultra in the Mac Pro and do that self handslap thing you do when you say something like “job done”?

Probably because Mac Pro sales are so rare that it isn't worth it.

 

[Riot Nrrrd]

I guess i also thought that Apple could open up the Mac Pro for add ons like the old days, but these days it’s all baked in or can be done via Thunderbolt 4/5.

Agreed. In 2011 I bought a used Mid 2010 Mac Pro cheese grater. Over the years I updated the RAM from 8 GB to 24 GB, I added a 480 GB SSD, I upgraded the DVD drive to a Blu-ray drive, I upgraded the video card to an AMD RADEON HD 7950 Sapphire Mac edition, and I added a USB 3.1 PCI-e card.

Even if the current M2 Ultra-based Mac Pro were affordable (it’s not), it just doesn’t feel like you can keep it for a decade and do the kinds of upgrades us Prosumers were able to do with the ol’ cheese grater cMPs. Without that kind of flexibility there’s no way I could justify buying one today, even if I could afford to (I can’t).

 

[Plerf]

Probably because Mac Pro sales are so rare that it isn't worth it.

The counterargument is maybe Mac Pro sales are so rare because it’s not a good value compared to the Studio.

My brother has a 2014 27" iMac which he has been trying to do exactly this. However it's easier said than done as software support is falling off due to Apple having dropped OS support on his system. He ran into issues running his tax software this year. Due to this he has resigned himself to buying a new Mac.

But which Mac should he buy? He likes the idea of the AIO but the screen size on the current iMac is leaving him underwhelmed. He's also considering a Mini with an external display but this configuration is not ideal for him.

He'd absolutely love to have a 27" M4 / M4 Pro iMac with which to replace his current one. Unfortunately that's not available and I don't understand why as the 27" iMac appears to have been very popular.

there certainly must be some sort of mount to slap a mini onto the bottom/back of a large display and thus creating a sort-of back door aesthetically questionable AIO.

Googling turned up this: https://www.etsy.com/listing/1809425516/apple-studio-display-mac-mini-m4-mount

 

[chucker23n1]

The counterargument is maybe Mac Pro sales are so rare because it’s not a good value compared to the Studio.

Yes, but "tossing the M3 Ultra in the Mac Pro" isn't going to fix that. Even with a lower price tag, the Mac Pro just isn't an attractive option right now unless you really, really need internal PCIe cards. It would require a different approach — either a more specialized SoC, or no SoC at all.

 

[Plerf]

Yes, but "tossing the M3 Ultra in the Mac Pro" isn't going to fix that. Even with a lower price tag, the Mac Pro just isn't an attractive option right now unless you really, really need internal PCIe cards. It would require a different approach — either a more specialized SoC, or no SoC at all.

Yeah. I suppose people buying Pros instead of Studios care more about the internal PCIe than maximum computing powah.

But still, for that money, they should be getting the very best Apple has to offer. Even if it’s not “profitable” to upgrade the Pro, they should do it anyway for goodwill.

Anyway, I’ll never be in the Ultra or even Max market and unlikely to ever return to desktop computering so it’s all just a wondering for me.

 

[chucker23n1]

Yeah. I suppose people buying Pros instead of Studios care more about the internal PCIe than maximum computing powah.

But still, for that money, they should be getting the very best Apple has to offer. Even if it’s not “profitable” to upgrade the Pro, they should do it anyway for goodwill.

Anyway, I’ll never be in the Ultra or even Max market and unlikely to ever return to desktop computering so it’s all just a wondering for me.

Yeah…

Rumor has it there's another SoC variant coming, and presumably that's either for the Mac Pro, or for whatever succeeds the Pro and/or Studio. I can't imagine Apple wants to stay with the awkward "the Pro is the Studio, but pricier and with PCIe slots" solution for too long. I don't think anyone is happy with that.

 

[Chancha]

I am still confident that an upper SoC has to come, if not WWDC then must be within the year. Otherwise it’d make too little sense not to chip drop M3 Ultra in the Mac Pro as well and call it a day now. Clear neglect on the flagship Pro desktop isn’t the bad press Apple wants again. Those higher ups who did the Trashcan apology roundtable are still running the company.

 

[obviouslogic]

Honestly, I am surprised that Apple has updated the Ultra variant every generation, so far. I just assumed it would be every other generation. The supposed performance jump from the Max to Ultra should have been great enough to sustain a large enough gap even over next generation Max chips.

The release of M3 Ultra though is rather odd, especially after everything else has been updated to the M4. I understand why Apple did it, but it is still weird and I kind of think the N3B process that Apple pushed for, while in the short term was great, in the long term it is turning out to be a mistake that Apple is having to deal with. Contractually, they may still owe production to TSMC, which is why the new iPad Air uses the M3 and one reason we have the M3 Ultra.

I doubt there will be an M4 Ultra, but I also think the M3 Ultra will end up being the shortest lived Ultra. I also believe Apple is moving away from UltraFusion and they’re developing a new SoC for both their high-end systems and their own custom servers. The lower end of which will eventually go in the Studio and the high end in the Mac Pro. Apple will introduce this SoC at WWDC and ship it in the Fall/Winter in the Mac Pro. Then next year, a binned version will replace the M3 Ultra in the Studio.

 

[woocintosh]

In the future, I think they might focus on enhancing the performance of the Max chip instead of the Ultra chip. If just one generation passes and the Max chip almost catches up to the Ultra chip, there might not be any need to keep making the Ultra chip. Maybe they could divide the Max chip into three tiers or something.

 

[peter2]

Just release a multi-CPU Mac Pro with multiple (at least dual) M4 Max chips. Ideally with extra RAM slots and support for PCIe GPUs.

 

[obviouslogic]

Keep in mind that came at a time when there was a Macintosh Performa 5420CD, not to be confused with the Macintosh Performance 5410CD, based on the Power Macintosh 5400. Am I supposed to get an LC, a Performa, or a Power Macintosh? Who knows! The poor salesperson certainly doesn't. Plus, Apple at the time relabeled and sold CD-ROM drives, scanners, printers, and even digital cameras. Why? Hard to say.

Just a nitpick here… These were not relabeled products. Yes, the core technology was from another company, but the products were Apple. (The LaserWriter used Canon engines, the QuickTake used image sensors from another company.) Calling them “relabeled” would be the same as calling an iPad, a relabeled LG/Samsung display.

 

[Velli]

I’m not sure what all the fuss is about. Enterprise level chips have never been about the latest and fastest cores, but about things like memory bandwidth. It was always the case with Intel that the latest i7/i9 etc was faster than Xeon in “normal” use cases like gaming. I think it makes perfect sense if Apple decides to only upgrade the Ultra every other generation, so we might still see an M5 Ultra. Given the rumours that Apple’s coming AI server farms will be running M5 gen chips, it would make a lot of sense.

Whether or not they choose this or to stop making Ultras altogether, if we don’t see a Mac Pro with M3 Ultra soon, I think Mac Pro is not long for this world.

 

[russell_314]

Yes, it is possible. But with Apple who knows if they will. A 27" M4 / M4 Pro iMac seems like a no brainer but here we are without one (or something similar).

This is true you don’t know if the rumors are correct. I have a feeling many of these people are just guessing or they get bad information. It’s possible Apple could have an iMac and some test lab and leakers take that as it’s going to market it’s really not

It's as if Apple goes out of their way to leave customers stranded.

I don’t see how anyone would be stranded. Granted you might not get exactly what you want, but Apple does make the Apple studio display that is compatible with both the mini and studio. While there may be some aesthetic differences, the functionality is the same. There is a big argument against the all in one computer because once the computer is out of date, then the monitor goes in the trash. This isn’t exactly ideal for the environment or consumers having to buy both items combined. I agree it looks better though.

Agreed. I have the 27” iMac with an i9 CPU. I just ordered a Mac Studio with M4 Max. That will last for years to come.

That is a good choice. It’s definitely going to be more powerful than what you have and more efficient.

I wonder who qualifies as a pro user anymore.

I don’t think Apple knows this answer. I love to call a lot of their products “Pro”. I think it’s become a marketing gimmick rather than a real description. I mean you think about it if you own something called pro it must be good, right? I want saw an advertisement for incontinence products and it was labeling them as “pro”. How long does it take before you become a pro at incontinence? 😂

I work with large arrays of x-ray spectra. The Studio will be awesome. Others working in geology or meteorology or other should also be happy. At what point do you step up to a small supercomputer instead of a desktop?

Well, if you’re doing this for work, then you are a pro user. I bet there are pro users that the mini would be perfect for. Someone like a professional accountant. Is the guy doing the books for a large corporation less of a pro because he doesn’t need more processing power? I think it’s all about the right tool for the job.

 

[chucker23n1]

Just a nitpick here… These were not relabeled products. Yes, the core technology was from another company, but the products were Apple. (The LaserWriter used Canon engines, the QuickTake used image sensors from another company.) Calling them “relabeled” would be the same as calling an iPad, a relabeled LG/Samsung display.

Sure, I was exaggerating, but surely the Apple-specific engineering effort in an iPad is way higher than it was in a StyleWriter. It was mostly "give it your own case design, and your own driver". With the iPad, it's their own SoC, their own selection of components, their whole own OS (including numerous first-party apps), etc.

 

[chucker23n1]

Just release a multi-CPU Mac Pro with multiple (at least dual) M4 Max chips. Ideally with extra RAM slots and support for PCIe GPUs.

I mean… the M4 Max is already a twelve-core CPU. Adding another

 

[obviouslogic]

As far as long-term success, the only Mac I'm currently unsure about is the Pro, but that isn't about customer confusion. Nobody is unsure "should I get the Pro instead?". They'll know when they need it. The Pro, however, has an uncertain future because so few people need it.

I think over the past decade Apple’s attention and support of their “Pro” customers has been somewhat chaotic. Currently I see the problem as having more to do with the transition to Apple Silicon and a huge misstep in their plans for it; that being the use of UltraFusion to create a 4x SoC variant. The rumors and proof were there for it being a thing, but I think they learned the hard way that it just wasn’t going to be feasible, either due to cost or performance issues.

Which means 3 years ago they had to start over. This left them shoving an Ultra 2 into the Mac Pro and calling it a day. I seriously doubt this was Apple’s game plan for the Pro. Especially after spending so much time concentrating on their Pro customers for the brief moment of time just before the transition; iMac Pro, really fast Xeon Pros., creating a Pro Workflow group within Apple, etc..

So I kind of think Apple ditched UltraFusion (after they already designed the M3) and started with a clean slate to create higher end SoCs. I think we’ll start to see that work either this year or next. With a high-end Studio, the Mac Pro and possibly their own custom servers, they may have the volume needed to justify creating a new high-end variant.