Who wants to run the Logic tracks benchmark on their M2/M3?

[aliensporebomb]

Can some of you report which benchmark test you are running to get these results?

The link had 3 items in it, two of them didn't seem to reflect what some of you were pulling up:

-Diva Test
-Logic Alchemy Test
-NewLogicBenchmarkTest2

Which one? How do we best run this to get best results?

 

[ori69]

Can some of you report which benchmark test you are running to get these results?

The link had 3 items in it, two of them didn't seem to reflect what some of you were pulling up:

-Diva Test
-Logic Alchemy Test
-NewLogicBenchmarkTest2

Which one? How do we best run this to get best results?

NewLogicBenchmarkTest2

 

[aliensporebomb]

Okay. After messing with this for a while:

M2 Max Mac Studio stock memory with 2 TB internal SSD and buffer size of 32 - I stopped after 246 tracks..

I went briefly to 250 but it wasn't reliable so backed off a little. I might back up a little more too if I'm on
the internal drive - external SSD via Firewire might let more happen.

Going to higher buffer sizes was a recipe for failure, I had to back WAY off to get actual functioning with higher
buffer sizes.

As TommesPommes said, M2 Max likes low buffer sizes.

My old iMac i7 4790 I'd get maybe 54 tracks? So this by far exceeds my old system.

 

[ori69]

MacBook Pro M4 Max 420 tracks.

 

[nathansz]

Okay. After messing with this for a while:

M2 Max Mac Studio stock memory with 2 TB internal SSD and buffer size of 32 - I stopped after 246 tracks..

I went briefly to 250 but it wasn't reliable so backed off a little. I might back up a little more too if I'm on
the internal drive - external SSD via Firewire might let more happen.

Going to higher buffer sizes was a recipe for failure, I had to back WAY off to get actual functioning with higher
buffer sizes.

As TommesPommes said, M2 Max likes low buffer sizes.

My old iMac i7 4790 I'd get maybe 54 tracks? So this by far exceeds my old system.

Something I’ve always found interesting about the new logic benchmark test is that any machine I’ve ever tried it on has a hard line

Given the same buffer size and all else being equal, if say 322 tracks works fine indefinitely and 323 tracks fails, then it’s repeatable over and over again

 

[krell100]

It's interesting that Logic only utilizes P cores, and doesn't seem to use E cores at all. Other DAWs (such as Reaper) use both P and E cores for much higher performance. I'd be interested to know why that is.

I guess this means if you're a Logic user then a Mac with a higher ratio of P to E cores is ideal, and total number of cores is less relevant than the total number of P cores.

 

[nathansz]

It's interesting that Logic only utilizes P cores, and doesn't seem to use E cores at all. Other DAWs (such as Reaper) use both P and E cores for much higher performance. I'd be interested to know why that is.

I guess this means if you're a Logic user then a Mac with a higher ratio of P to E cores is ideal, and total number of cores is less relevant than the total number of P cores.

A Mac with a higher ratio of p to e cores is probably always ideal

Thats why I returned my m3 pro as soon as the m4 pro specs came out

….


really though, its not going to be an issue either way for logic

People get caught up on specs and benchmarks but pretty much any modern computer is good enough for pretty much anything outside of gaming

Maybe you can bounce out a bit faster

I was just in a studio that was all m2 Mac minis running pro tools

compute power was a non issue

 

[krell100]

@nathansz Would you say that this still holds true for those using very demanding plugins such as Diva and Pigments?

It seems that a lot of benchmark tests focus on track count but I've found that a single track with a lot of intensive plugins in the chain can bring even a high spec Mac/PC to it's knees.

 

[nathansz]

@nathansz Would you say that this still holds true for those using very demanding plugins such as Diva and Pigments?

Good question


I’m not a very heavy vst user so Im the wrong person to ask

I’ll demo at home with alchemy or what have you but then it’s generally all real synths in the studio

 

[streetfunk]

It seems that a lot of benchmark tests focus on track count

i feel same.
It´s totally not doing justice for all form of usecases or workflows.

but I've found that a single track with a lot of intensive plugins in the chain can bring even a high spec Mac/PC to it's knees.

agreed.