Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
They should have put the M4 Max version out now, with a M4 Ultra option coming later this year or something. The M3 Ultra makes no sense.

Performance wise it fits nicely as a step above the M4 Max. And to be honest it sits right about where I assumed a new monolithic M4 Ultra would’ve hit. GB6 multi-core scores should reach 30,000+ and Metal scores should definitely hit close, if not above, 300,000.

CPU performance on the M3 Max was just under the performance of the M2 Ultra, so doubling the M3 Max is a huge leap.

Apple has recently stated they will not make an Ultra for every generation (even though they have so far). So, I would take that to mean, there probably won’t be an M4 Ultra.

However, since they didn’t also upgrade the Mac Pro, it makes me wonder if they don’t have something else up their sleeve for WWDC? If Apple is going to start using their own Apple Silicon based system in their servers, maybe they designed a new architecture for them and it will be used in the Mac Pro?
 
  • Like
Reactions: Allen_Wentz
If Apple had the M3 Ultra in the on-deck circle, why didn't they just release an M3 Max/M3 Ultra Mac Studio last year?

Because the M3 Ultra clearly wasn't ready at that point. They likely intended to release it last year, but only just got yields where they needed them.

There would be no point giving the base Studio an M3 Max now though, when the M4 Max has been shipping for months. Doing so would have been a massive middle finger to their Studio customers (and uncompetitive with PC competition, especially for GPU). In any case, I seem to remember the M3 generation being on an expensive process node, and Apple being keen to move to M4.
 
So were stuck with a 5yr old monitor?

Unless you're willing to accept the indignity of a non-Apple monitor, and pick up a ASUS PA27JCV ProArt 5K Display. Heck, you could get two for the price of an Apple display.
 
So were stuck with a 5yr old monitor?
There are several new displays with the same 27" 5k Panel. Asus PA27JCV, BenQ PD2730S, ViewSonic VP2788. I'm waiting for reviews for the ViewSonic.
 
Apple pimping out their users for a $20B check from Google somewhat says otherwise. Apple gets that huge check from Google for all the work the customers put into the system and Apple hands how much of that back directly to the customers?

$400/TB SSD capacity prices are 'customer first'?

Performance wise it fits nicely as a step above the M4 Max. And to be honest it sits right about where I assumed a new monolithic M4 Ultra would’ve hit. GB6 multi-core scores should reach 30,000+ and Metal scores should definitely hit close, if not above, 300,000.

CPU performance on the M3 Max was just under the performance of the M2 Ultra, so doubling the M3 Max is a huge leap.

Apple has recently stated they will not make an Ultra for every generation (even though they have so far). So, I would take that to mean, there probably won’t be an M4 Ultra.

However, since they didn’t also upgrade the Mac Pro, it makes me wonder if they don’t have something else up their sleeve for WWDC? If Apple is going to start using their own Apple Silicon based system in their servers, maybe they designed a new architecture for them and it will be used in the Mac Pro?
This:
"...something else up their sleeve for WWDC? If Apple is going to start using their own Apple Silicon based system in their servers, maybe they designed a new architecture for them and it will be used in the Mac Pro?"
 
That's not how it works. SKUs of Apple Silicon chips are named after their CPU core generation.
Why? Says who? You know why the 5th generation x86 processor from Intel was called Pentium instead of 586? Because a court ruled you can't trademark sequential numbers. Maybe Apple is doing something like that. Nothing forces Apple to do it the same way just because they have done it that way in the past. I will freely admit it is the most likely answer, but that doesn't make it the only one.
True, but that's the only difference. And as an IO technology, is literally peripheral.
It is the only difference you can tell based on the order page. Unless you have the M3 Ultra already and tested everything, you don't know this yet. TB5 might literally be a peripheral, but it is literally an important one for the target audience, who would be a lot less interested in buying without it. Everyone complains when Apple doesn't support the absolute latest spec (USB, HDMI, TB, and so on), so they added it this time and get "but that's the only thing."
Don't make out poor old Apple is in a no-win situation. If they had called an M3-based product an M4 (easily verified with e.g. Geekbench), customers would obviously complain about the deceit. If people complain there is no M4 Ultra, it's because Apple failed to deliver it.

If they were able to deliver M4 Ultra, they obviously would have done. But clearly, something went wrong and it was either delay the Studio further or release now. Given it was still on M2, Apple sensibly decided that an update took priority. Most buyers likely get the Max anyway, and the M3 Ultra is still a big step up on the M2.
Where did you see the Geekbench results for the M3 Ultra? I was looking before I even made the post, and no M3 Ultra data was present. I am not saying it is no-win for Apple. As a front runner for world's most valuable company, they clearly are winning. But there are some people in this very thread who just want to complain about Apple not doing enough.

I have an M1 Ultra Studio, and have been wanting to upgrade. What I wanted to see come out this week was an M4 Ultra Mac Pro. Clearly, it isn't Christmas morning for me. If they had rolled out an M3 Ultra Pro, I would have ordered last night. An M3 Ultra Studio has me waffling. Based on M2 Ultra numbers, I think the Multicore, Metal, and OpenCL performance will be a notable improvement, and I really want hardware raytracing, but I wanted it in a Mac Pro. I think I might sit on this for a bit and see if the computer I am wanting is going to show up later in the year.
 
Why? Says who? You know why the 5th generation x86 processor from Intel was called Pentium instead of 586? Because a court ruled you can't trademark sequential numbers. Maybe Apple is doing something like that. Nothing forces Apple to do it the same way just because they have done it that way in the past. I will freely admit it is the most likely answer, but that doesn't make it the only one.

It is the only difference you can tell based on the order page. Unless you have the M3 Ultra already and tested everything, you don't know this yet. TB5 might literally be a peripheral, but it is literally an important one for the target audience, who would be a lot less interested in buying without it. Everyone complains when Apple doesn't support the absolute latest spec (USB, HDMI, TB, and so on), so they added it this time and get "but that's the only thing."

Where did you see the Geekbench results for the M3 Ultra? I was looking before I even made the post, and no M3 Ultra data was present. I am not saying it is no-win for Apple. As a front runner for world's most valuable company, they clearly are winning. But there are some people in this very thread who just want to complain about Apple not doing enough.

I have an M1 Ultra Studio, and have been wanting to upgrade. What I wanted to see come out this week was an M4 Ultra Mac Pro. Clearly, it isn't Christmas morning for me. If they had rolled out an M3 Ultra Pro, I would have ordered last night. An M3 Ultra Studio has me waffling. Based on M2 Ultra numbers, I think the Multicore, Metal, and OpenCL performance will be a notable improvement, and I really want hardware raytracing, but I wanted it in a Mac Pro. I think I might sit on this for a bit and see if the computer I am wanting is going to show up later in the year.
IMO odds are very high that when we see a MP it will be an M5 with its primary design aimed at serving AI. An M5 MP is an exciting proposition. Apple might make it into a swiss army knife of the high end.

Adding up to 500 GB RAM to the newly announced M3 Ultra makes it a real beast for many applications. I doubt if an M4 Ultra will ever be coming.
 
Where did you see the Geekbench results for the M3 Ultra? I was looking before I even made the post, and no M3 Ultra data was present. I am not saying it is no-win for Apple. As a front runner for world's most valuable company, they clearly are winning. But there are some people in this very thread who just want to complain about Apple not doing enough.
I'd say it's fairly easy to extrapolate. For the "base Ultra" spec, just take the 14 core M3 max and double the multi-core score. Subtract maybe 5% for losses in the interconnect and bob's your uncle.
 
Why? Says who?

Erm, Apple. Given the precedent set to date, calling the M4 Ultra the M3 Ultra would not only be confusing, but self-defeating and therefore pointless. It would be like selling a "1 carat" ring with 2 carat's worth of diamonds.

It is the only difference you can tell based on the order page.

Right, I'm sure Apple have packed the M3 Ultra full of goodies they're not telling us about. Perhaps as a nice surprise for purchasers?

Everyone complains when Apple doesn't support the absolute latest spec (USB, HDMI, TB, and so on), so they added it this time and get "but that's the only thing."

But it is the only thing. Sure, TB5 is better than TB4, and welcome. What people are disappointed about is that Apple are using the older architecture in an expensive, top of the range computer. With the M5 out in ~7 months.

Where did you see the Geekbench results for the M3 Ultra?

My point was there's no need to decap the SoC or anything. Utilities like Geekbench can directly identify the CPU / GPU cores, or simply reveal their presence based on comparing e.g. ST performance with existing SoCs. So if Apple had marketed the M3 Ultra as an M4, they'd have got found out very quickly, with accompanying blowback. So that wouldn't have been an option.

I have an M1 Ultra Studio, and have been wanting to upgrade. What I wanted to see come out this week was an M4 Ultra Mac Pro. Clearly, it isn't Christmas morning for me. If they had rolled out an M3 Ultra Pro, I would have ordered last night. An M3 Ultra Studio has me waffling. Based on M2 Ultra numbers, I think the Multicore, Metal, and OpenCL performance will be a notable improvement, and I really want hardware raytracing, but I wanted it in a Mac Pro. I think I might sit on this for a bit and see if the computer I am wanting is going to show up later in the year.

If you'd have ordered an M3U MP same day as it was announced, it sounds like you would be happy with its performance. Given fitting one to the Mac Pro would only have required a revised logic board, it doesn't look like Apple are particularly interested in updating it, though.

I'm not sure I'd wait for a Mac Pro. I can't see them making two Ultra's in one year, with one exclusive to a machine that only sells around 200k a year. And Apple's SoCs hardly have oodles of spare PCIe, so you're probably better off with TB peripherals or a TB5 expansion case.
 
The first thing I did when this news came out was to check to see if I was still within the return period for the Mac mini I purchased a month ago (answer: nope). I'd been reading rumors that the M4 Mac Studio wouldn't be out until mid-year and didnt' want to wait that long.
This is the problem with their current product release strategy. Nobody knows what or when updates are coming. Makes it impossible to plan ahead, which is especially problematic if you have a business.
 
  • Like
Reactions: jido and AlumaMac
after reading these ultra M3 threads, I wouldn't be surprised if this is the last year of the Studio Ultra. I think they may have originally thought the Ultra Studio with peripherals would maybe replace the mac pro, but then decided to put it all out there. I also think there are many things that go into the release decisions that we don't even know about, so anythings possible. But clearly there is way more negative reaction to this than positive. I ordered one, but wonder if I'm way in the minority.
 
Performance wise it fits nicely as a step above the M4 Max. And to be honest it sits right about where I assumed a new monolithic M4 Ultra would’ve hit. GB6 multi-core scores should reach 30,000+ and Metal scores should definitely hit close, if not above, 300,000.

CPU performance on the M3 Max was just under the performance of the M2 Ultra, so doubling the M3 Max is a huge leap.

Apple has recently stated they will not make an Ultra for every generation (even though they have so far). So, I would take that to mean, there probably won’t be an M4 Ultra.

However, since they didn’t also upgrade the Mac Pro, it makes me wonder if they don’t have something else up their sleeve for WWDC? If Apple is going to start using their own Apple Silicon based system in their servers, maybe they designed a new architecture for them and it will be used in the Mac Pro?
I may be missing something, but unless the M3 Ultra is something entirely new, I think the M4 Max will perform very closely, if not better in some areas, to the M3 Ultra - certainly not worth doubling the price ($2K) without special GPU needs. I compared each generation's increase and found the following on Geekbench (numbers are rounded, and M3 Ultra specs are based on the average increase percentage between the M1 Max to M1 Ultra and M2 Max to M2 Ultra. Also note that the M2 Max to M2 Ultra did not have the same level of increase as the M1's did, a declining trend but with only two data points). I also used the base CPU of each, as Apple charges a ton more for slightly upgraded chips and I don't think most people buy them.

It appears that with the exception of Metal, the M4 Max will perform very similarly to

M1 Max to M1 Ultra performance increase:
  1. Single core: 2400, basically the same between Max and Ultra
  2. Multi-core: 13K to 18K (38% increase)
  3. OpenCL: 62K to 95K (53% incease)
  4. Metal: 106K to 162K (52% increase)

M2 Max to M2 Ultra performance increase:
  1. Single core: 2800, basically the same again
  2. Multi-core: 15K to 21K (40% increase)
  3. OpenCL: 88K to 116K (31% increase)
  4. Metal: 146K to 203K (39% increase)
M3 Max to M3 Ultra estimated performance increase based on M1 and M2 increases:
  1. Single core: 3100
  2. Multi-core: 19K – 26K (39% avg. est. increase)
  3. OpenCL: 78K – 110K (41% avg. est. increase)
  4. Metal: 155K – 225K (45% avg. est. increase)
M4 Max performance:
  1. Single core: 3900
  2. Multi-core: 23K (26K with 16C M4 Max)
  3. OpenCL: 100K
  4. Metal: 160K
If there was an M4 Ultra, using the same estimated performance increases above for the M3 Ultra, you'd get:
  1. Single core: 3900
  2. Multi-core: 32K
  3. OpenCL: 141K
  4. Metal: 232K
 
Last edited:
after reading these ultra M3 threads, I wouldn't be surprised if this is the last year of the Studio Ultra. I think they may have originally thought the Ultra Studio with peripherals would maybe replace the mac pro, but then decided to put it all out there. I also think there are many things that go into the release decisions that we don't even know about, so anythings possible. But clearly there is way more negative reaction to this than positive. I ordered one, but wonder if I'm way in the minority.
See this is what I was referring to with my comment that Apple’s attitude to the pro market is highly questionable to where I am not sure if the Studio or Pro will even remain. Studio won’t exist with just the Mac chip.
 
I don’t give a damn about their share price or shareholders.


What you care individually has little to no impact to Apple.


If their strategy is putting shareholders before customers they won’t remain the world’s richest company.

Their priority is to service customers to generate the most money for them. By keeping those customers happy and buying they continue to generate more revenue with growth making shareholders happy.
 
I don’t see any reason why the world’s richest company couldn’t handle the release of several Macs at once.
The thing about the “why can’t the world’s richest company (fill in the blank)” argument is that everyone uses it. And if Apple actually did everything that everyone who uses that argument demands, Apple would no longer be the world’s richest company. They might even be the world’s poorest judging by some of the extreme demands I’ve seen and the likely sheer number of them.
Some people seem to equate “richest” with “limitless”. What they don’t realize is that at Apple’s scale, ALL the numbers get bigger, both the good and bad. There are absolutely limits and will always be in our physical world. Any company that acts otherwise acts irresponsibly.
Steve Jobs said Apple is like the world’s largest startup company. That commitment to efficiency is a big reason for their financial success. Some people are essentially asking Apple to throw that strategy out the window.
 
All the gate-keeping of BTO options behind specific chips and core configurations.

You used to be able to get the base Studio and upgrade to 64GB RAM with that standard processor. Now it's 36GB, and the 48GB or 64GB options require a different M4 Max with two extra cores, effectively increasing the price significantly. Locking certain SSD sizes to specific chips doesn't even have a technical reason to it, it's just a margin move by Tim.
Well, the more BTO options, the more it costs Apple in operations and logistics. I don’t know if you count this as a technical reason though since it still ultimately comes back to money.
 
  • Like
Reactions: picpicmac
That’s why they will never succeed in the corporate market and still only have 10% market share in computers after 50 years in business.

I've worked in various tech roles in the enterprise market for a while for multiple large companies and no offense but this is one of the most ridiculous takes I've read on these forums to date for a number of reasons which I could spend all day on. Heck even Apple conceded any chance at that market with their own marketing with those Im a PC commercials stating PCs are better at 'spreadsheets' though that would be a simpleton overview of the reasons.



Without the iPhone Apple would be worthless.

Well duh. Thats the general idea as they are not a computer(laptop/desktop) company first like in the old days. Did you ever wonder why Steve renamed the company from Apple Computer, Inc. to Apple Inc. in 2007 when their market cap was 74 billion and then the iPhone comes out and 5 years they are worth 500 billion and now we are in the trillions?

Take a wild guess. As a computer company back then focused on Mac OSX they would always be limited. Mac OSX era couldn't even beat Windows for certain entertainment like gaming let alone trying to beat them in the enterprise market which was actually impossible. Apple saw what things like the iPod and iTunes could do for the company and thus the strategy changed and it paid off better than anyone could have dreamed of.
 
[Apple] and still only have 10% market share in computers after 50 years in business.
Ok, here's a question: roll back 50 years to 1975: How many computer companies back then are still in the business of selling computers today?

Answer: basically none. IBM today sells services.

But, if you go down just one year, to 49 years, do you know who's next?

Apple.

And your claim of 10% is just wrong. There are only four major players in the small computer market, each has between 20% and 25% of the market: Lenovo, Dell, HP, and Apple.

Apple has the least volume of the four... but the most revenues.

Corporate buyers are concerned with bulk costs and support costs. For example, my local health care giant has thousands of Lenovo computers and monitors in their many offices. No one really cares about what brand is on the monitor. Some corporate buyer likely just shopped a purchase order and Lenovo gave them the best deal.


That's how it works. No one in those offices care if the screen is 1440p or 2160p. No one cares if there is an AMD or Intel processor inside. No one cares if there is a seperate GPU or if the graphics is integrated in the main processor.

Those are things that internet geeks argue over.

In that big world of computer buyers, by which you want to judge Apple, no one cares about the model number on the ICs in their tools.

Only those on the internet who want to argue over model numbers - it's a "3" and not a "4" - care about those things.
 
(3) Since a Studio is architecturally sort of like a "Maxi" Mac mini, I wonder if it will be subject to Rearportgate™, where USB devices have been seen randomly disconnecting from the rear Thunderbolt ports on the M4-based Mac mini.
If you read the article you linked, it seems the issue affected some who tried those little USB-c to USB-A adapters.

The Studio already has two USB-A ports, so the luddites can use those.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.