What's new

Apple Announces Macbook Pro with M2 Max and 96gb RAM.

I like my Ultra a lot.


I've been very happy with my M1 MacBook Air for live work using MainStage. I just keep it out of the sun when performing outdoors.

The Mac Studio has a much larger case, and the Ultra has copper or something - it's heavier than the Max because of its extra heat sinking.

I think the word "Studio" in the name isn't only marketing, it's the application they're targeting: creative elites like all of us here. :)
 
I'm really happy that I bought the M1 Pro MBP at the end of last year and did not wait for this. The hefty price bump is certainly not worth the performance gain for me.
I have a slight regret not going for the 16" and 4TB of storage though, but I'll manage.
 
“MacBook Pro 16 with M2 Max is hotter and louder than the M1 Max machine was. While running PugetBench, the M2 Max machine’s fans were so loud that people across the office were coming over to see what was going on. I would not have wanted to be typing at that time — the top of the keyboard was toasty. I only saw this during PugetBench — it was not at all loud or hot during everyday Chrome usage — but the M1 Max machine, by contrast, was astonishingly silent and cool throughout my testing period.”
 

Where's the test that relates to Audio?

I get that there is concern re: heat with the M2 Laptops. OTOH, it appears the tests that have pushed the heat envelope are GPU-related.

There is no audio application of which I am aware that will even warm up a GPU. As such, I'm wondering whether the expressed concerns may be misplaced, at least for the usage I'd expect from VI-Control members (unless they are also video editors.)
 

Where's the test that relates to Audio?

I get that there is concern re: heat with the M2 Laptops. OTOH, it appears the tests that have pushed the heat envelope are GPU-related.

There is no audio application of which I am aware that will even warm up a GPU. As such, I'm wondering whether the expressed concerns may be misplaced, at least for the usage I'd expect from VI-Control members (unless they are also video editors.)
Indeed, which is why I'm thinking the M1Max would be the better choice, especially since Logic Pro benchmarks show identical numbers.
 
Just in. Again, same numbers for Logic Pro between M1Max and M2Max but the latter goes up to 109 Celsius in graphics applications, which doesn’t bode well for longevity.


2:20 (worse speakers)
3:30 (smaller heat block)
14:40 (109C erratic and louder)
18:00 (identical Logic Pro benchmark)
 
Last edited:
e.
1B. Most DAWs allow you to disable instruments not in use, freeing up resources when not used, but keeping them immediately available once enabled.

im curious on this one with m2 pro max (or m1 ultra).

I turn off/on tracks (along w ram) but if i have for example, the strings section in a summing folder and aciddentally press the main folder with 20 tracks, there goes my afternoon. I have mac pro 5,1 and it takes forever to load a bunch. for each individually its about a few seconds. Not too bad but still a little cumbersome when trying to work.

Anyone using this type of template with the new m series and see a big difference in response? not sure how much faster it is with better single/multi core of the m2.
 
Just in. Again, same numbers for Logic Pro between M1Max and M2Max but the latter goes up to 109 Celsius in graphics applications, which doesn’t bode well for longevity.


2:20 (worse speakers)
3:30 (smaller heat block)
14:40 (109C erratic and louder)
18:00 (identical Logic Pro benchmark)

for me the main difference is being able to add more ram for the same class of cpu. Which leads to me be ok in waiting for the m2 series Mac studio so i can have a pro max with 96gb of ram. (hopefully) and save some money. The ultra is crazy fast but ill be more than fine with the pro max except that the limit is 64gb of ram.

Starting to see now a pattern with the m series like we are seeing with iPhones and its incremental upgrades, price points and feature upgrades. So yearly releases that get a little better but most poeple will buy every 3-5 years.
 
The similar M1/M2 performance on the Logic test is interesting and a bit of a head scratcher; I wonder if Logic updates may impact this as time rolls along.

One consideration not contemplated in that test: Single Core Performance.

Some audio tasks can run up against single-core limit, in particular some CPU-intensive VI patches and/or multi-plugin stacks on a single track. The M2 single-core performance advantage could come into play in these situations.
 
Single core difference is still quite limited (compared to multi core). Here’s hoping this gets better with the M3 (3nm) chips. And that overheating should be gone too.
 
Why does Logic not take advantage of multiple cores?

Do other DAW's take advantage of this? It seems bass-ackwards that Logic would limit itself to single core
 
It's only an issue for any instruments that are being played live. That is, when a VI is being played from an external keyboard, that instrument and all the effects in the path to the output will be using a single core. It can be maddening when you have a VI that simply won't play without breaking up because of a CPU core overload when you see all the rest of the cores are practically taking a nap, there's so little activity. But, I don't know enough about what's involved to render VIs in real time to know if this is just the way it is or if there's a workaround that Logic doesn't take advantage of.
 
Why does Logic not take advantage of multiple cores?

Do other DAW's take advantage of this? It seems bass-ackwards that Logic would limit itself to single core
It uses all the cores, or rather: you can instruct Logic to work in either of these ways:

Logic cores.png
I just made a quick test with Modern Scoring Strings, having five Kontakt instances open at the same time, with unique instruments – all with four dynamic legato layers and automation of dynamics and vibrato. This was on an 2020 i7 3,8 GHz iMac, and it played all these tracks without using more than between 50% and 75% of one single core.

Performance Meter 5 instruments.png
I may be wrong about this, but I'll give it a go: I think it's only if one puts so much stuff on one single track that one core cannot handle it that Logic can't switch to using multiple cores for one single track. But since this test shows that one core can handle 20 dynamic layers of legato/vibrato and a total of 8 mic positions activated, the situation isn't really bad. There are iPads out there with better performance than this iMac (which has the best single core performance of all Intel Macs), but of course YMMW.

When playing live (four out of the 5 instruments from the playback tracks, copied to a separate track), the meters looks like this:

25 instruments playback + 5 live.png
If I try with five MSS instruments on one track instead of four (in addition to the 25 playback instruments spread over 5 tracks), the meter in the core used for live playing hits the roof. All this is with a 256 buffer in Logic.

The performance in the M2 and M1 Macs is of course better than this.


Mac mini (2023)
Apple M2 Pro @ 3.5 GHz (12 cores)
1943

Mac mini (2023)
Apple M2 @ 3.5 GHz (8 cores)
1933

https://browser.geekbench.com/macs/mac-mini-2023-10c-cpu (Mac mini (2023))
Apple M2 Pro @ 3.5 GHz (10 cores)
1929

https://browser.geekbench.com/macs/macbook-pro-13-inch-2022 (MacBook Pro (13-inch, 2022))
Apple M2 @ 3.5 GHz (8 cores)
1901

MacBook Air (2022)
Apple M2 @ 3.5 GHz (8 cores)
1892

Mac Studio
Apple M1 Max @ 3.2 GHz (10 cores)
1756

Mac Studio
Apple M1 Ultra @ 3.2 GHz (20 cores)
1754

MacBook Pro (16-inch, 2021)
Apple M1 Max @ 3.2 GHz (10 cores)
1745

MacBook Pro (14-inch, 2021)
Apple M1 Max @ 3.2 GHz (10 cores)
1745

MacBook Pro (16-inch, 2021)
Apple M1 Pro @ 3.2 GHz (10 cores)
1742

MacBook Pro (14-inch, 2021)
Apple M1 Pro @ 3.2 GHz (10 cores)
1737

MacBook Pro (14-inch, 2021)
Apple M1 Pro @ 3.2 GHz (8 cores)
1731

iMac (24-inch Mid 2021)
Apple M1 @ 3.2 GHz (8 cores)
1720

Mac mini (Late 2020)
Apple M1 @ 3.2 GHz (8 cores)
1715

iMac (24-inch Mid 2021)
Apple M1 @ 3.2 GHz (8 cores)
1715

MacBook Pro (13-inch Late 2020)
Apple M1 @ 3.2 GHz (8 cores)
1708

MacBook Air (Late 2020)
Apple M1 @ 3.2 GHz (8 cores)
1707

https://browser.geekbench.com/macs/imac-27-inch-retina-mid-2020-intel-core-i7-10700k-3-8-ghz-8-cores (<b><span style="color: rgb(226, 80, 65)">iMac (27-inch Retina Mid 2020)</span></b>)
Intel Core i7-10700K @ 3.8 GHz (8 cores)
1250

iMac (27-inch Retina Mid 2020)
Intel Core i9-10910 @ 3.6 GHz (10 cores)
1241
 
Last edited:
It uses all the cores, or rather: you can instruct Logic to work in either of these ways:

Logic cores.png
I just made a quick test with Modern Scoring Strings, having five Kontakt instances open at the same time, with unique instruments – all with four dynamic legato layers and automation of dynamics and vibrato. This was on an 2020 i7 3,8 GHz iMac, and it played all these tracks without using more than between 50% and 75% of one single core.

Performance Meter 5 instruments.png
I may be wrong about this, but I'll give it a go: I think it's only if one puts so much stuff on one single track that one core cannot handle it that Logic can't switch to using multiple cores for one single track. But since this test shows that one core can handle 20 dynamic layers of legato/vibrato and a total of 8 mic positions activated, the situation isn't really bad. There are iPads out there with better performance than this iMac (which has the best single core performance of all Intel Macs), but of course YMMW.

When playing live (four out of the 5 instruments from the playback tracks, copied to a separate track), the meters looks like this:

25 instruments playback + 5 live.png
If I try with five MSS instruments on one track instead of four (in addition to the 25 playback instruments spread over 5 tracks), the meter in the core used for live playing hits the roof. All this is with a 256 buffer in Logic.

The performance in the M2 and M1 Macs is of course better than this.


Mac mini (2023)
Apple M2 Pro @ 3.5 GHz (12 cores)
1943

Mac mini (2023)
Apple M2 @ 3.5 GHz (8 cores)
1933

https://browser.geekbench.com/macs/mac-mini-2023-10c-cpu (Mac mini (2023))
Apple M2 Pro @ 3.5 GHz (10 cores)
1929

https://browser.geekbench.com/macs/macbook-pro-13-inch-2022 (MacBook Pro (13-inch, 2022))
Apple M2 @ 3.5 GHz (8 cores)
1901

MacBook Air (2022)
Apple M2 @ 3.5 GHz (8 cores)
1892

Mac Studio
Apple M1 Max @ 3.2 GHz (10 cores)
1756

Mac Studio
Apple M1 Ultra @ 3.2 GHz (20 cores)
1754

MacBook Pro (16-inch, 2021)
Apple M1 Max @ 3.2 GHz (10 cores)
1745

MacBook Pro (14-inch, 2021)
Apple M1 Max @ 3.2 GHz (10 cores)
1745

MacBook Pro (16-inch, 2021)
Apple M1 Pro @ 3.2 GHz (10 cores)
1742

MacBook Pro (14-inch, 2021)
Apple M1 Pro @ 3.2 GHz (10 cores)
1737

MacBook Pro (14-inch, 2021)
Apple M1 Pro @ 3.2 GHz (8 cores)
1731

iMac (24-inch Mid 2021)
Apple M1 @ 3.2 GHz (8 cores)
1720

Mac mini (Late 2020)
Apple M1 @ 3.2 GHz (8 cores)
1715

iMac (24-inch Mid 2021)
Apple M1 @ 3.2 GHz (8 cores)
1715

MacBook Pro (13-inch Late 2020)
Apple M1 @ 3.2 GHz (8 cores)
1708

MacBook Air (Late 2020)
Apple M1 @ 3.2 GHz (8 cores)
1707

https://browser.geekbench.com/macs/imac-27-inch-retina-mid-2020-intel-core-i7-10700k-3-8-ghz-8-cores (<b><span style="color: rgb(226, 80, 65)">iMac (27-inch Retina Mid 2020)</span></b>)
Intel Core i7-10700K @ 3.8 GHz (8 cores)
1250

iMac (27-inch Retina Mid 2020)
Intel Core i9-10910 @ 3.6 GHz (10 cores)
1241
It does take advantage of multiple cores.
Then why all the talk of "single core" performance? I am genuinely asking - thank you
 
Top Bottom