Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Your conclusion is flawed because your premises are flawed. M4 was never intended as an Ultra platform; it lacks the necessary connection points.
You're speaking out of ignorance. You have no information on whether the M4 never had such connection points or not.

Even if the currently-released M4 Max chips do lack them (as indicated by Apple's press release), it's entirely possible Apple developed variant M4 Max prototypes with such connection points, designed just for the M4 Ultra, but abandoned the M4 Ultra project when they realized they couldn't get it working in time.

Indeed, that may be the case for the Max chips in the M3 Ultra (that they are a different variant than the already-released M3 Max chips). We already have evidence that they differ at least somewhat from the original M3 Max chips, since the former have TB5 controllers, while the latter have TB4.

Someone who lacks the ablility to consider such options should not presume to insultingly lecture others on the quality of their logic.
 
Last edited:
Very confusing. M4Max is worse than M3 Ultra? Apple needs to try and make better sense of their chips naming scheme for our sakes.

I think if people stop and think about this for a second they’ll realize this is common practice in all industries for ages.

Example:
Car manufacturer releases a model. It has three engine options.
Same manufacturer releases an updated version of that model, with three engine options.
The fastest of the old is faster than the slowest of the new.

And at the end of the day the average user doesn’t care, nor do they need to care.
 
You're speaking out of ignorance. You have no information on whether the M4 has such connection points or not. Just like many asserted the M3 lacked such connection points.

And even if the currently-released M4 Max chips do lack them, it's entirely possible they had variant M4 Max prototypes with such connection points, designed just for the M4 Ultra, but abandoned the M4 Ultra when they realized they couldn't get it working in time.

Some who lacks the ablility to consider such options should not presume to lecture others on the quality of their logic.
M3 having UltraFusion connection or not was just rumors (well we learned today that they do have it and didn't build a M3 Ultra from scratch but indeed combined 2 M3 Max chips) but today Apple confirmed to the press that M4 does not have this UltraFusion thing.

So maybe they'll skip the M4 Ultra and come up with a M4 Extreme for the Mac Pro update ?
 
Multiply the M3 Max laptop by 2 versus the single M4 Max score. That is a pretty good proxy for the parallel workloads.
If primarily concerned with just single threaded stuff then there little about Max vs Ultra that makes a difference.
Indeed, if one is primarily concerned about single-threaded workflows, then the M4 Max is superior to the M3 Ultra for processing speed (unless they decide to boost the max clock on the Ultra specifically).

However, while this would be wasteful, in theory someone doing single-threaded calculations on very large matrices might need an Ultra just for the much larger max memory size. [More plausibly, they might buy an Ultra if they needed it for mixed ST and MT work, and needed the large RAM for both.]
 
Last edited:
macOS has a horrible time with docks and hubs.
I'm using a cheap Ugreen four-port USB hub right now, attached to a Mac, to type this. The keyboard is attached to it, and a Logitech dongle (for their trackball) is attached to it. Works fine, has always worked fine.
 
Apple confirmed that no commerically-released M4 Max chips have/will have the UltraFusion bridge, which tells us there will be no M4 Ultra.
The M1 Ultra was quickly (by Apple flagship product standards) replaced by the M2 Ultra.

Which of course makes us wonder if the M3 Ultra will be replaced by the M5 Ultra next spring.
 
I couldn't get to the Apple Store fast enough to order the Studio M4 Max.
I just wish the M4 Max could do more than 128GB of RAM. Otherwise, I think you're right. And if one can buy at the discount price the M4 Max Studio is quite attractive.
 
M4 ultra in the Mac Pro or - skip it entirely for a partial redesign for the M5 that would lead the pack? Will be interesting to see what's what in the upcoming months.
 
Yes, Apple confirmed that no commerically-released M4 Max chips have/will have the UltraFusion bridge, which tells us there will be no M4 Ultra. But that doesn't rule out the possiblity they were prototyping an M4 Max variant with the UF bridge, in the hope of creating an M4 Ultra.

Possible versus probable are two different things. If this M3 Max+ used for Ultras is cut off from the flow of M4 Max (laptop/plain) versions in the most of the Mac Studios sold the volume on these alternative , augmented Max+ dies, then this is going to a relatively , very , very low volume. Even if Apple used them for Private Cloud Compute that would just get it out of the 'crazy' low zone into just very low.

Super low volume is a problem. A different die mask is somewhat offset by less expensive plain Max , but it is a vastly smaller user base to amortize the additional costs that don't get covered.

Pretty good chance that the Mac Studio Ultra and Mac Pro Ultra have to combine forces to get to anywhere near decent amount of volume. The MBP 14/16" Max's were subsibdizing a large portion that would be different. But just 'half' of the Studio or just the MP by itself , that is not going to likely be economically viable.


Unless you know what happened internally at Apple during their development process, you don't know know what their original plans were for the M4 Ultra—whether they were hoping to release it alongside the M4 Max Studio and failed, or whether they never intended to release it, period.

the M4 Max in a Mac Studio had a bit of a problem because the M4 Max is in same ballpark as M2 Ultra. So keeping the Studio on M2 Ultra for longer period of time was likely to lead to increased fratricide problems.

If the MBP 14/16" are not soaking up all the M4 Max SoC why would apple sit on inventory so that the Studio and Mac Pro can launch together. They don't have to be hyper tightly coupled. It isn't about development. it is also about chip supply. That is one reason why the much higher volume MBA is only getting M4 now.
 
M4 ultra in the Mac Pro or - skip it entirely for a partial redesign for the M5 that would lead the pack? Will be interesting to see what's what in the upcoming months.

If Ultras are not coming every generation then this "lead the pack" thing is an illusion. It is just a sporadic bragging rights goal. it isn't a fill real user needs goal.

If the UltraFusion was 'missing' from the M3 Max plain/laptop version., then Apple has clear opportunities to add more Mac Pro feature enhancing I/O to the M3 Max+ using for the Ultra. If there is PCI-e v5 dual xPCI-e v8 there for the PCI-e switch backhaul in the M3 Ultra ..... Apple should skip that waiting for the M5 something? That makes no sense.

If Thunderbolt was tweaked up for M3 Max+ , then it wouldn't be all that hard to uptick the PCI-e back haul in addition.

What was holding Mac Pro back is being hyper coupled to the MBP 14/16" Max die. If Apple is loosing that a bit then there is room to improve in dimensions other than the baseline arch version number.

The Ultras are just likely going to be on separate iteration cycles. Even more so if Apple increasing the decoupling from the much higher unit volumes of the MBP 14/16".
 
So maybe they'll skip the M4 Ultra and come up with a M4 Extreme for the Mac Pro update ?

The pricing of the M3 Ultra points toward there not being an Extreme coming any time soon.
The require minumal RAM pricing , plus the addition development costs are going to push prices to where almost nobody is going to buy it.

If Apple has decoupled the building block die for the Ultra from the MBP 14/16" ( a die incrementally different than the one used in the relatively much higher volume laptops ) then the Ultra itself is on a slippery slope. The Extreme is like off the cliff on costs/affordability.

Apple needs to sell a lot more Ultra Mac Pro's far more than some even smaller Extreme models. Lower numbers (for much higher prices ) isn't going to help the Mac Pro.
 
Upgraded my base model Mac Studio M1 with Mac Studio M3 Ultra base model for $3999. I thought the upgrade processor would be nice, but not for $1500. I will be perfectly content with the base model for the next 4-5 years.
 
  • Like
Reactions: Defender Stargate
I ended up going with the M4 Max. Came from an M1 Ultra, and if I'm honest with myself I don't need the extra power or RAM for $2k more. Thankfully they both have DisplayPort 2.1, so I won't need hacks anymore to get the full resolution on a Samsung G9 57".
 
  • Like
Reactions: phrehdd
M3 Ultra lol. They are trickling these CPU updates. M4 Ultra would clearly blow things away but with how clear of a lead M-series chips are they don't need to release an M4 Ultra now.
 
Lots of angst in this thread. I can only say that as a 2018 Mac Mini Intel owner, I couldn't get to the Apple Store fast enough to order the Studio M4 Max. What a machine.
Same. I didn't like the new Mini footprint, and I sure as heck didn't need an M4 Ultra. This is the perfect next step for me.
 
The Mac Pro is in a weird place now:

M2 Ultra, 64GB RAM, 1TB storage - $6999

vs Mac Studio:

M3 Ultra, 96GB RAM, 1TB storage - $3999

Heck, you could even option for 256GB RAM on the Studio and it would still be cheaper than the Pro!

You got to have some really specific needs to justify the Mac Pro?
There was a bulk sale 6 months agos in US, some dealers offloaded Base M2 Mac Pro's for 3500, even I didn't play retail price where it wasn't on sale officially. 128G with wheels (that was the one available at the time), it was 1500 cheaper than the base model. For me it came out much cheaper than a M2 Ultra Mac Studio. 2019 Mac Pro's are still more expensive here. It's hilarious.

I can wait for 2 generations easy, unless Apple removes support for M1/M2 fast. I find it unlikely.
 
I don’t see any reason why the world’s richest company couldn’t handle the release of several Macs at once.



If you bothered reading their 10-ks over the years like investors you would understand their business better.

Spoiler: Apple put priority towards items that produce most revenue and push stock price up with headlines. It's priority is not appeasing niche desktop users because they are the richest company in the world.
 
  • Like
Reactions: picpicmac
If you bothered reading their 10-ks over the years like investors you would understand their business better.

Spoiler: Apple put priority towards items that produce most revenue and push stock price up with headlines. It's priority is not appeasing niche desktop users because they are the richest company in the world.
I don’t give a damn about their share price or shareholders. If their strategy is putting shareholders before customers they won’t remain the world’s richest company.
 
  • Like
Reactions: krell100
Say what you want about Intel, AMD and Nvidia. At least these manufactures are not confusing with their releases per. year

Up until the M2 cycle everything was uniform with the launch
- Spring = MacBook Air with Base M-Series
- Fall [Oct] = Macbook Pros with M-Pro and M-Max Series
- WWDC following year = Previous year M-Ultra

Now it's just all kinds of messed.
Please tell me you are joking.
i3, i5, i7, i9
but an i5 14500 versus an i7 13K series, and AMD makes all that look intuitive.
 
  • Like
Reactions: picpicmac
The dumbest marketing strategy of all time? Why on Earth would they make an M4 Ultra, then call it an M3 Ultra? Obviously not the case.

Perhaps there was no space left on the M4 die for an UF connector.
Well, technically, it is the 3rd "Ultra" chip, even if it is based on M8. So there is that.
It has thunderbolt 5 and the M3 pro didn't.
If they HAD called it M4 Ultra, those same people would be complaining because there was no M3 ultra, even if the new one was based on 'next gen M5 chips.'
 
If their strategy is putting shareholders before customers they won’t remain the world’s richest company.
Apple do put its customers first.

Here is the pecking order for Apple customers:
1) iPhone buyers;
2) Service (e.g., Apple Music) buyers;
3) MacBook buyers;
4) iPad buyers;
5) Watch buyers;
6) Apple TV buyers;
7) AirPod buyers;
8) AVP buyers (could be bumped up);
9) AirTag buyers;
10) Mac desktop buyers;
11) Home Pod buyers.

So if you're whining about the MacStudio it's because you are low on the list of priorities.
 
The real question is, how much faster is M3 Ultra compared to M4 Max
I am looking at Geekbench, and it is looking like it will be notable.
In multicore, the M4 Max beats the M2 Ultra beats the M3 Max
In Open CL and Metal, the M2 Ultra reigns supreme.
In Single Core, even the M2 Max beats the M2 Ultra, and the M4 Max is way faster, even 20% faster than an M3 Max. Only, who would buy one of these for single core applications (well, about half this thread now that I have mentioned this, but I mean people who would actually buy it, not post about how they were going to but now 'will never buy Apple again every year').
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.