But with a desktop, power efficiency is not really important since you are always plugged in and the Studio Ultra has a very impressive cooling system with plenty of headroom.
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.
Is “Max” Mac Studio the worst option?
- Thread starter cocoua
- Start date
- Sort by reaction score
But with a desktop, power efficiency is not really important since you are always plugged in and the Studio Ultra has a very impressive cooling system with plenty of headroom.
Power efficiency is important pretty much all the time in this day and age. If you look google for "M2 Ultra thermal throttling" there are plenty of examples. Everything is a balancing act / tradeoff, but I don't see how literally zero e-cores makes any sense from what we know today. Just on the basis of computers spending most of their time idling, you'd want at LEAST two.
Maybe the ratio of P to E will just be moved toward P cores on the Ultra? And maybe the Ultra could be clocked slightly higher too to really see some gains.
I always find the 2 to 8 ratio of E-Cores to P-Cores on the M1 Max Studio to be good. M1 Ultra getting 4 to 16 is the same ratio but 4 already feels excessive.
On the M2 Max and M2 Ultra there are 4 to 8 and then 8 to 16 which is where it starts looking like it needs adjustment. 8 E-Cores has no business being in a semi-workstation class machine, you probably only need 2-4 if not just 1 of them when you are not pushing the performance envelope.
On the M2 Max and M2 Ultra there are 4 to 8 and then 8 to 16 which is where it starts looking like it needs adjustment. 8 E-Cores has no business being in a semi-workstation class machine, you probably only need 2-4 if not just 1 of them when you are not pushing the performance envelope.
For low-stress workloads, you could have the active performance cores clock themselves down and the inactive cores be at idle to minimize the thermal envelope.
If Apple adds efficiency cores to the M3 ULTRA, then they would use one four-core cluster (as found on the M3 and M3 MAX).
I have not seen the die shots, but presuming the performance cores are arranged in six-core clusters (one for the M3 PRO and two for the M3 MAX), I would expect Apple to offer a four and six cluster option, allowing for 24 or 36 pCores. And the "M3 EXTREME", if it exists, could offer eight or more clusters (for 48+ pCores).
NO. It's about the performance that each individual CHOOSE to have for their own personal convenience to do whatever they need to do on their Mac Studio!!! I have the M2 Max Mac Studio since August 2024 and have no issues with its performance at all. If I wanted better, then I would have chosen the M2 Ultra or Mac Pro, especially that I came from using a Mac Pro in the first place. Just like when you go out and buy a car, are you going to get what you can for the best financial performance for your bucks or the worst?
+1. I’m assuming that if Apple decides to go with zero E cores on the next gen Ultra, then they will have a way to keep it from Throttling (especially in the Studio). If this is the case and someone wants the extra power and gives little or no weight to efficiency then go for the Ultra. It’s up to the individual.
If this is true, it will be very disapointing that the Mac Studio M4 is released a whole year after iPad M4, when people is buying M5...
www.macrumors.com
Many apps depend on single core, and buying a M3 6000USD screen/battery/keyboard-less machine this June, slower than an iPad in some task would be like the old Intel days where you should sacrifice cores per single core performance.
The alleged advantage you would get paying such a high price wouldnt match your real advantage in the market.
To be clear, buying a Mac Studio M2 today has no sense at all (unless obviously, you get a >20% discount or you really really need it for work). But for one just thinking in updating systems, the Mac Studio only has sense buying in at it release date (before next Macbook Pro MPro/Max would be release 6 months later to match price and performance)
The real thing here would be to release the Ultra the first one. The MSMax together with the ultra would be too beautiful to be true, and Apple wouldnt do that never as it would cannibalize MBP16" sales., But the Ultra is too far from the general public and is the Mac's flagship model. It belong to be the first.
M4 Macs Are Expected to Launch in This Order Starting Later This Year
Bloomberg's Mark Gurman recently reported that the first Macs with M4 series chips will be released later this year, with more models to follow...
www.macrumors.com
Many apps depend on single core, and buying a M3 6000USD screen/battery/keyboard-less machine this June, slower than an iPad in some task would be like the old Intel days where you should sacrifice cores per single core performance.
The alleged advantage you would get paying such a high price wouldnt match your real advantage in the market.
To be clear, buying a Mac Studio M2 today has no sense at all (unless obviously, you get a >20% discount or you really really need it for work). But for one just thinking in updating systems, the Mac Studio only has sense buying in at it release date (before next Macbook Pro MPro/Max would be release 6 months later to match price and performance)
The real thing here would be to release the Ultra the first one. The MSMax together with the ultra would be too beautiful to be true, and Apple wouldnt do that never as it would cannibalize MBP16" sales., But the Ultra is too far from the general public and is the Mac's flagship model. It belong to be the first.
Last edited:
yeah, the 8 E-cores totally feels as a design flaw waiting to be fixed, lets see this June what they have for us
