Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

mac57mac57

macrumors 6502a
Original poster
Aug 2, 2024
514
310
Myrtle Beach, SC
I currently have an M1 Max based Mac Studio and am very pleased with it. I have however been eagerly awaiting the release of the M4 Max powered Mac Studio, expecting a huge bump in performance. However, looking at Apple's own performance information on their web site, I see this:

M4 Max vs.jpg



If I am reading this correctly, the new M4 Max equipped Mac Studio is not even twice as fast as an M1 Max equipped Mac Studio; by my math, it is only about 60% faster than its M1 Max predecessor.

If you want a truly significant performance bump, you have to equip your new Mac Studio with an M3 Ultra, and pay quite a bit more for the priveledge.

Based on these numbers, I am going to stay with my M1 Max equipped Mac Studio. I will upgrade when the "Max" processor in a new Mac Studio delivers at least 2x the performance.

I *am* surprised that going from M1 to M4 does not do *more* than 2x, not less than!
 
  • Like
Reactions: thatguyshan
Based on these numbers, I am going to stay with my M1 Max equipped Mac Studio. I will upgrade when the "Max" processor in a new Mac Studio delivers at least 2x the performance.
I'd wait until some hard benchmarks come out. It also depends on what software you specifically are using. The M4 Max is a very fast machine, especially if the software you use is not massively multiprocessor-enabled. Single-core speeds of 2369 vs 4054 (per Geekbench's library of tests) is not an insignificant boost. And multicore speeds are 12197 vs 25913 (which is 2x as fast).

But it all depends on whether that speed improvement is worth the difference in cost. For some it is, for some it isn't.
 
  • Like
Reactions: thatguyshan
That one Apple test is only concerned about PS task that happens to use the neural engine, hardly anyone’s main photo workflow.

Check this petapixel comprehensive review:
The focus is M3 Ultra but M4 Max data is included, as with other older Macs, some tests included your M1 Max.
(I am assuming you work with photography, since you posted that PS test)
 
Blender M4 Max 40 c GPU vs M1 Max 32 c GPU

Schermafbeelding 2025-03-15 om 02.00.15.png
Schermafbeelding 2025-03-15 om 02.00.26.png


Cinebench M4 Max 40 c GPU vs M1 Max 32 c GPU

Schermafbeelding 2025-03-15 om 02.03.06.png
Schermafbeelding 2025-03-15 om 02.03.28.png


Novabench M4 Max 40 c GPU vs M1 Max 32 c GPU

Schermafbeelding 2025-03-15 om 02.09.40.png
Schermafbeelding 2025-03-15 om 02.09.59.png
 
Last edited:
  • Like
Reactions: picpicmac
My personal work experience between my loaded M1 Ultra and my loaded M4 Max MBP, is there is not even a comparison. The new M4 Max is radically faster, in every single regard. It's the first time, in all my years using an Apple product (i.e. since their inception) where the gains have been so dramatic, in only a few years.
 
Geekbench 6
Studio M1 Max 32 GPU
Studio M4 Max 40 GPU

Parallels Desktop 20 with Windows 11. M4 Max vs M1 Max
Paralllels 20 Windows 11 Passmark M4.png
Parallels 20.1 Win 11 Passmark 11 21H2.png
 
Last edited:
Max Tech exists to find ways to exploit the YouTube algorithms. That said, all one has to do is look at the benchmark tests and see that the M4 Max is faster than the M2 Ultra in nearly all uses, the exception being video exports.
 
@dougnash, I remember when the very first PowerPC Macs came out, running System 7. We got them at work, and we were all so excited... there had been SO much hype about the PowerPC RISC architecture and it's performance benefits. But then, all the Mac code was still 68K and that fast, fast, fast PowerPC ended up for the most part emulating 68K and the net result was ... about the same! No real improvement.

Nice to see that Apple is actually moving the yardsticks with new product these days!
 
  • Like
Reactions: EdwardC
@dougnash, I remember when the very first PowerPC Macs came out, running System 7. We got them at work, and we were all so excited... there had been SO much hype about the PowerPC RISC architecture and it's performance benefits. But then, all the Mac code was still 68K and that fast, fast, fast PowerPC ended up for the most part emulating 68K and the net result was ... about the same! No real improvement.

Nice to see that Apple is actually moving the yardsticks with new product these days!
And don't forget the good ol' days, of having to restart the entire machine, dozens of times a day; up until the revelation that was OSX.
 
And don't forget the good ol' days, of having to restart the entire machine, dozens of times a day; up until the revelation that was OSX.
Back then, the solution to ANY problem (according to inexperienced bozos in web forums) was to "Rebuild the Desktop" and "Zap the PRAM" – neither of which would have any affect on the problem at hand, but it sure made some people sound like geniuses. Good ol' days, indeed!
 
... and "Repair Permissions", which was also often (up through Tiger, at least) touted as the cure-all for ANY problem!

I realize that I am now mixing generations (Mac OS, and Mac OS X) but it just fits the discussion!
 
... and "Repair Permissions", which was also often (up through Tiger, at least) touted as the cure-all for ANY problem!

I realize that I am now mixing generations (Mac OS, and Mac OS X) but it just fits the discussion!
Oh my gosh, I forgot about that "cure-all" myth. That was another classic. Those people never did figure out that the act of restarting their Mac was the actual cure, not all the voodoo they did right before that.

And with that... I'm off to defrag my SSD drive! 😆
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.