I have the M2 Max Studio Base model.
I could've made the session more efficient if I needed to, because I didn't really need 9 Batteries or 3 EZ Drummers for example. I could've consolidated some of those together.
So for me, in the past I have always gone for the max number of CPU's with speed greater than 2.8GHz (when turbo'ed or not), max RAM, max read/write speeds, and that seems to have worked fine for DAW work (composing / writing, initial mixing, stem production etc). My workstation was also used by our studio video editor for rendering, hence the need for internal SSD speeds etc - although that was when the team and I were all based from the same studio location. These days, the team is much much smaller (downsized due to COVID), and generally work from home or a client studio, so the editing / rendering need on my workstation is no longer so great, but having learned from our build experiences of 2017/18/19, I tend to follow the same principles.
Perhaps I missed it in an earlier post? Where or how are you storing your VI samples? Are you using 2.5" SSD drives in external enclosure's, or NVME's in a thunderblade external type of enclosure?
Currently just using the internal drive. But I'm more of a synth head and don't use lots of VI samples. My concern is mostly CPU usage, not so much RAM and storage.
You're using this, right?
Ok..so the internal HD. Those are usually screaming fast, but no good for me. I always bench mark with the blackmagic app and use it for the occasional troubleshooting and yes, I do get 300-400MB/s write and 500MB/s read speeds with all of my drives that are very well used. Mostly Samsung 850pro's and 870 EVO's. As you said, VI's are little bits of data so nothing like the 1GB or 5GB sustained read/write that blackmagic performs. Greatly improved load times would be super....but I'm looking at around $5500 for a 32TB thunderblade which is almost the computer cost LOL. I probably would stick with my TB3 Thunderbay MIni's at first, benchmark the heck out of them and then I can always switch to the NVME thunderblade if I think I should be experiencing more performance. I just never know how much of the session loading time for VEPro/Digital Performer is the CPU having to think about each of the VI's to load vs the actual load time for loading the samples into RAM.
Probably just because they’re different sessions with different instrument and processing plug-ins.
This excerpt is haunting me now that I’m choosing an m1 vs m2 for Logic Pro
I guess the source is this article:
Im still not getting it. :(
I wouldn't worry about a technicality in the logic board. Yeah of course the M2 is better but that's expected because it's the new gen.
Good to know . The m1 still haven’t come down in price as much as I thought. (Yet) so both seem a close callI wouldn't worry about a technicality in the logic board. Yeah of course the M2 is better but that's expected because it's the new gen.
Your decision is basically if you want to pay for more GPU performance and 192GB of RAM or not
The biggest improvement of the M2 Ultra is in the GPU which is massive. In CPU it's like 20-30% better, which is nice, but not life changing.
The M1 Ultra kicks ass. Specially for audio.
I don't see anything more about it either. My hunch is that DAWs have been distributing workload across cores for years, so if the cores are available, a DAW will use them. So, the more cores the merrier for our work. But, I know very little about benchmarks. I would be interested to see a test project with the CPU Usage Window of Activity Monitor open and tested on an M1 Max, an M1 Ultra, an M2 Max, and an M2 Ultra.
This is similar to my workload and the M2 Max should be able to handle it easily, which is why I decided to go with that.
Thanks for the reply! That was super helpful.
