Little help explaining current speed differences between chips and laptops/studio/mini etc ?

[Dahkoht]

Figure this forum would be good for getting folks with experience in both Apple and Windows hardware to be able to enlighten me a little on this.

Background I'm a sys engineer for 20+ years , built tons of PC's , have had some Macbook Pro's in the past I loved etc , been away from Apple Ecosystem and about to get either a studio or macbook pro , maybe Imac , depending on what gets announced Monday night. It's going to be a secondary machine running some Windows 11 ARM virtual plus just me getting back into learning the ecosystem overall too. For both work and learning ,I've got a separate windows gaming rig and work laptop if need too.

Anyway my question is can anyone maybe explain the major differences in the Apple silicon both generation wise and macbook pro vs studio use etc.

For example , I know a 3080 vs 4060 Nvidia card is night and day even if the 4x series is "newer" , that a 10900k vs an 11400 is the same comparison and so on. Also that a 3080 in a desktop vs a laptop in the windows world is night and day etc too.

So in the Apple Silicon world is there any difference between an M2 Max in a macbook vs a studio ?

Is an M2 Max likely to still be better overall than an upcoming M3 base , and so on ,

Just would appreciate if someone familiar with both sides could enlighten a mainly Windows engineer who is familiar with Apple but been a while what the current state of Apple Silicon is using a comparison.

 

[UnknownSouljer]

Figure this forum would be good for getting folks with experience in both Apple and Windows hardware to be able to enlighten me a little on this.

Background I'm a sys engineer for 20+ years , built tons of PC's , have had some Macbook Pro's in the past I loved etc , been away from Apple Ecosystem and about to get either a studio or macbook pro , maybe Imac , depending on what gets announced Monday night. It's going to be a secondary machine running some Windows 11 ARM virtual plus just me getting back into learning the ecosystem overall too. For both work and learning ,I've got a separate windows gaming rig and work laptop if need too.

Anyway my question is can anyone maybe explain the major differences in the Apple silicon both generation wise and macbook pro vs studio use etc.

For example , I know a 3080 vs 4060 Nvidia card is night and day even if the 4x series is "newer" , that a 10900k vs an 11400 is the same comparison and so on. Also that a 3080 in a desktop vs a laptop in the windows world is night and day etc too.

I'm not going to break every piece of this down. Mostly because literally all of this information is available from a quick Google search. It's not hard to find out all of this stuff in about 5 minutes.

In fact here is this chart breaking down the entire M2 lineup in terms of core counts: https://en.wikipedia.org/wiki/Apple_M2
A 2 second Wiki search.

So in the Apple Silicon world is there any difference between an M2 Max in a macbook vs a studio ?

The short answer is: no. One difference that does occur is cooling, where-in the Studios don't throttle "as much" due to a larger cooling solution. But the chips are the same.

It's the area that makes their laptops so good, and their desktops feel a bit underwhelming; as Apple doesn't currently have an ultra-high wattage chip that is designed specifically for desktop.
The Ultra is desktop only, and it is "twice as fast" as a Max (it's two Max chips essentially), but cost wise it doesn't make sense for most unless you're in a field that benefits from the accelerators. And it's obviously a pro machine for people making money with their computers.

Is an M2 Max likely to still be better overall than an upcoming M3 base , and so on ,

The very short version is the whole product stack is incredibly well organized and thought out. And unlike the PC side actually makes sense numerically (referring both to the CPU and GPU market where naming can change on a whim). And each generation they are more or less sticking to the formula of more of the same, but faster.

From bottom to top every aspect of the chips in the stack simply grow as you go up the stack.

The base is "the smallest" in all ways. In physical die size, in core count, in GPU core count, and sometimes lacking other features like the media engines (which was changed from M1 to M2).

-M1 and M2 were both 8 core machines 4P + 4C cores. and 8-10 core GPUs.
-The Pro adds more CPU and GPU cores and has specifically more performance cores. 6P + 4C cores. up to 8P + 4C. The Pro is twice the size of base silicon. Which is a significant amount of transistors.
-The Max adds more. 8P + 4C cores. Most of the change here on M1 and M2 is in amount of GPU cores. Up to 38. The Max adds another half Pro size again. It's a much larger chip.
-The Ultra is literally two Max chips attached to one another with a 5TB/s link. So 24 cores and 76 GPU cores.

There are other differences too. The amount of RAM that can be addressed by each system increases as you go up the product stack. The current Max chip can address 96GB of RAM and the Ultra 192GB of RAM. Whereas base can address only up to 24GB of RAM.

So knowing this information, it's obvious that a base M3 will never be faster than a previous Pro chip. Because Apple is likely to always have core counts in both CPU and GPU be the major differentiator. So it would take a massive performance increase in single core on both CPU/GPU to eclipse a previous higher level machine.

In fact my M1 Max I expect will not be eclipsed by a base level machine likely for 3+ years (or 5 total). It actually may be longer than that. No one expects an Intel i3 to ever touch an i9 without probably 5+ years difference, in any metric. The only way that it will have that kind of dramatic shift is if they decide to up the CPU/GPU core counts of lower end machines and increase the size of the silicon that they use for lower end machines. Which they might do. But I don't expect that the base machines will exceed 10 cores any time soon.
Whereas at both the top and bottom on Apple there will likely be a 20% (minimum) jump generation to generation.