But damn, M4s on iPads and M2s on the Studio (not even M3s). It's damaging for Apple's reputation as a "pro" company at the very least (not arguing with you, just thinking aloud)...
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.
Mac Studio Speculation [Merged]
- Thread starter darkus
- Start date
- Sort by reaction score
It really is bizarre that a $600 M4 iPad has a significantly faster single core than the $4k M2 Ultra Studio or the $7K M2 Ultra Mac Pro.
Pretty sure they're more than $600. I priced one on the Apple website (1 TB model) and it was comfortably in base model Mac Studio territory...
Considering what the last event revealed about the M4: how much faster will an M4 Max Studio be over the current M2 Max Studio in base config? And how much faster will the GPU be in the upcoming M4 Max Studio?
PS this is the current base config:
PS this is the current base config:
I doubt that there will be any new M3 device.
This is mostly due to the fact that TSMCs N3B was not what everyone hoped for. Problems all the way, low yield rates with wavers. So why the iPad Pro got the M4 in N3E first? Because the sales are rather low compared to most other devices from Apple, means its better suited for a controlled start of N3E production.
Yields with N3E are way better.
And here we are why the M4 had to be redesigned, chip designs on N3B are not compatible with N3E.
So it wasnt because of AI whatever pressure why the M4 was now introduced, its the costs, which are significantly lower with N3E.
So whats the next device to go M4?
I bet it will the the iMac which is still on M2.
Mac Studio might be, but i would not expect this before fall.
As the next TSMS itteration N3P will be just a shrinking, dont expect the same huge performance benefits as from M3 to M4.
This all also means, dont expect the same same chip strategy with iPhones as in the last years.
From a cost perspective it doesnt make any sense for apple to produce the old architecture chips for cheaper phones. So they will go all A18 there.
This is mostly due to the fact that TSMCs N3B was not what everyone hoped for. Problems all the way, low yield rates with wavers. So why the iPad Pro got the M4 in N3E first? Because the sales are rather low compared to most other devices from Apple, means its better suited for a controlled start of N3E production.
Yields with N3E are way better.
And here we are why the M4 had to be redesigned, chip designs on N3B are not compatible with N3E.
So it wasnt because of AI whatever pressure why the M4 was now introduced, its the costs, which are significantly lower with N3E.
So whats the next device to go M4?
I bet it will the the iMac which is still on M2.
Mac Studio might be, but i would not expect this before fall.
As the next TSMS itteration N3P will be just a shrinking, dont expect the same huge performance benefits as from M3 to M4.
This all also means, dont expect the same same chip strategy with iPhones as in the last years.
From a cost perspective it doesnt make any sense for apple to produce the old architecture chips for cheaper phones. So they will go all A18 there.
Yep, pretty much.
People believing in M3 anything now are not up to date with the info so they are still stuck with previous pattern. The node process changed everything. M4 onwards now.
The only exception (totally illogical though) would be if Apple designed M3 for N3E but that would make zero sense on pretty much all levels. So no, no M3 from now on.
Makes sense. Base M4 Max Mac Studio could have 18 CPU Cores and 40-48 GPU cores, making it perform slightly better than the M2 Ultra and still consume less than 20watts of electricity for many tasks.
But... will it support Thunderbolt 5?
There never was an M2 iMac. The latest iMac is the M3 iMac, and it was the successor to the M1 iMac
The latest update from Gurman has arrived.
If accurate, the 2nd half '25 M4 Studio and Mac Pro are reiterated.
I suppose WWDC in a few weeks will spill the beans.
Or not.
www.macrumors.com
If accurate, the 2nd half '25 M4 Studio and Mac Pro are reiterated.
I suppose WWDC in a few weeks will spill the beans.
Or not.
Gurman: No New Mac Studio and Mac Pro Until Mid-2025
Apple will not refresh the Mac Studio and Mac Pro with next-generation high-end chips until the middle of 2025, according to Bloomberg's Mark...
www.macrumors.com
oh dearThe latest update from Gurman has arrived.
If accurate, the 2nd half '25 M4 Studio and Mac Pro are reiterated.
I suppose WWDC in a few weeks will spill the beans.
Or not.
Gurman: No New Mac Studio and Mac Pro Until Mid-2025
Apple will not refresh the Mac Studio and Mac Pro with next-generation high-end chips until the middle of 2025, according to Bloomberg's Mark...
it makes zero sense. Why give us Raytracing and Mesh etc. when the computers that need it the most are still on M2. What is Apple smoking?
To make things even weirder there are SUPPOSEDLY M4 Mac minis coming this year...
N3B yields are likely no where near as 'horrible' as folks on. 1.5 after volume production there should have been improvements.
Pretty good chance this is little to do with fab processes and far more to do with just generally relatively higher costs for much bigger dies. Couple that to the relatively much lower run-rate volume for the Ultra (or bigger) class of systems it gets worse.
There is also no place else for the Ultra to go. There is no "hand me down" product to shovel them into. That isn't a path to 'churning' the biggest SoC package and 'throwing it away' every 12-14 months. Nobody in the rest of the large package business does that ( Intel , Nvidia , AMD , etc. )
Pretty good chance the largest die is going to get onto a M(n) , M(n+2) , M(n+4 ), etc. pattern. So M2 -> M4 --> M6 .
Apple relatively quickly dumping the M2 Max out of the the MBP 14/16" likely just made things worse for a transition off of M2 for the rest of the products using that building block ( Studio and Mac Pro). Now they have to shoulder a larger share of covering the investment/R&D/etc.
iPhone SoCs come in and the phone gets sold for 3 years. In no way shape or from doesn't Apple completely drop the latest phone SoC in 12 months. It is sold for at least 4 years via other 'hand me down' routes through plain, entry iPad , AppleTV , etc.
What is relavent is not just when it comes to market for the first time. When it goes out (gets dropped) is a very significant issue also. The Max class die is 400+ mm^2 chip. Dumping that in the 'trash can' after just 12 months of retail would be very expensive.
The other issue is that Apple is reportedly not going to split the iPhone SoC across fab processes this year. The 'plain' and Pro iPhone are reportedly going to A18 (and A18 Pro? ). That is a demand bubble that could squeeze out M-series for a relatively new fab process that has longer 'bake' times.
Also because the M3 wouldn't work with the iPad Pro screen. Apple's display controller has been behind the curve for a while (e.g., ignoring DisplayPort 2.x ) . They chose M4 to merge that in.
Mini / Mini Pro can also 'sop up' M2 gen SoCs for bulk of 2024. ( again the rapid dump from the MBP's ; MBP 13" was 2nd best selling system. ).
Pretty good chance Apple is still felling some urgency in MBP space so those iterate out also in the Fall. (and M4 , M4 Pro available for Mini).
The hang up for the Studio is possibly hung up on the Ultra. And the Ultra couples the Studio and Mac Pro together. N3E has incrementally better yields , but it also tosses some density to by that. ( all the caches/SRAM get bigger. And biggest Apple M-series dies have the biggest cache area. There will be area increase creep there. They can work on trade-offs; smaller logic for larger caches. ). IF the "Ultra" dies creep up larger than the reticle limit then the packaging tech they used previously ( InFO-LSI ) doesn't work so well. CoWoS packaging tech has been swallowed up by the AI mega package hysteria ( huge backlogs ). If Apple has to switch packaging or go mega-monolithic that will probably be a time hiccup (if not originally on the roadmap).
But if you are talking big dies that are on edge of busting the reticle limit... shrinking might be necessary. The Max die was already skirting the limit when paired up. Apple designs are abnormally cache heavy. The compromise that N3E makes is impactful trade-off ( it was not 'free' fix. )
Why would that be 'weird' ? The Mini is squatting on M2 ... and the M4 is shipping already. So why wouldn't Apple use it this year?
As soon as the M4 Pro shows up then have both sides to do a update bump. The MBA M3 just came out so it likely isn't getting bumped in 2024. So the #1 selling system isn't going to 'consume all of the M4" supply. The plain M4 MBP sells at a higher price than MBA so it isn't a volume competitor the MBA is.
Apple has often let the Mac Pro drift in "Rip van Winkle" mode. The MP didn't update until after 'about two years for the transition. " They drifted on the iMac Pro. The iMac 27" got bumped last in the transition.
As long as the Studio and Mac Pro are coupled to somewhat exotic 3D packaging technology, it isn't likely on the 'hyper fast track'.
I cant see that Apple really doesnt update their most pro devices for 2 years (if 2025). In the case of the Mac Studio even longer. I also cant see a M3 in any device, just dosent make any sense from cost perspective.
As there are rumors that the Ultra wont be just the Max 2 times like in past, but a dedicated new chip which than can be doubled again into an Extreme version for the MacPro there are possibilities i see that we might see a Studio Update but without a Ultra Version, which might come later with a Mac pro update. And especially as these Pro devices dont sell soooo much, it wont overstretch production capacity at this point.
But we will see soon.
As there are rumors that the Ultra wont be just the Max 2 times like in past, but a dedicated new chip which than can be doubled again into an Extreme version for the MacPro there are possibilities i see that we might see a Studio Update but without a Ultra Version, which might come later with a Mac pro update. And especially as these Pro devices dont sell soooo much, it wont overstretch production capacity at this point.
But we will see soon.
Do you think the price of the Studio will increase with the M4? I've got quite a bit saved now for a new computer, a bit longer and I can get a Studio but would rather wait for an M4. I was looking at the MacBook Pro's and thinking maybe wait for that to get M4 and OLED but at £3600 it's a lot more money, like about a grand more and I bet the price goes up with an OLED screen.
The M2 Studio is more expensive than the M1 version here in Canada, so it probably depends on your local market. I’m expecting the next one to go up even more as the US dollar is so strong.
Don't get OLED for a computer. You will totally get burn in unless you baby your machine and use it few hours a day only.
OLED is still flawed tech. Burn in wasn't solved yet. Its reduced but its more prone to happen on a computer that is on for 10+ hours a day everyday.
Unless you use computer for just few hours a day then you will be pissed off to spend so much money and then have to deal with this 'built in' flaw
OLED is still flawed tech. Burn in wasn't solved yet. Its reduced but its more prone to happen on a computer that is on for 10+ hours a day everyday.
Unless you use computer for just few hours a day then you will be pissed off to spend so much money and then have to deal with this 'built in' flaw
Don't know where you get your info from, but they more or less have solved OLED burn in. I've got an OLED TV from 2019 and have used it a LOT and it doesn't have a trace of burn in, thanks to the tech it has built in to prevent it, and that's from 2019.
Yes they may not last as long as LCD. But it's not like they don't last years. But if you like anything you want a good quality screen that lasts you have to pay for it.
Exactly and with tandem OLED this is further reduced as each panel can cut on how bright it is and by this further reduce the aging process
Burn in is not solved. Burn in is reduced.
With computer being on for most people for 12+ hours a day with static stuff you will get burn in.
Don't know where you get your info from but thats the reality with OLED.
There are ways to reduce the risk but the risk is still there and will be for a while.
With computer being on for most people for 12+ hours a day with static stuff you will get burn in.
Don't know where you get your info from but thats the reality with OLED.
There are ways to reduce the risk but the risk is still there and will be for a while.
Do you know LCD screens also use to suffer from burn in? Some quite badly? But guess what the tech improved, and now every computer by default has an LCD panel and you hardly ever hear them having burn in issues.
OLED is exactly the same, I've played hours and hours and hours of games on my Sony OLED TV and even game HUD's on screen almost that entire time, I've never had any burn in issues.
These days OLED's really don't suffer burn in and OLED monitors are now becoming more and more widespread.
Longevity And Burn-In Investigation: 10-Month Results
It has been about ten months since we started our long-term accelerated longevity and burn-in test, and our 100+ TVs and monitors have cumulated over 500,000 hours of runtime.
www.rtings.com
"As for burn-in, well, unfortunately, it seems like burn-in is still an issue, but newer panels are clearly better than older generations, and it seems like most people shouldn't worry about it with varied content."
You really need to educate yourself if you think BURN IN is not an issue. It still is and will be for a while especially for content with static elements (computer has tons of them)
So, if you have varied content then you will probably be fine as the tech improved there. If you have CNN or any static content on for hours and hours then you will get OLED.
Lets not live in denial here!