HCMarkus
TL;DR
You will love the new M1 Mac. I'm extremely happy with my M1 Mac Studio.
Apple Announces Macbook Pro with M2 Max and 96gb RAM.
- Thread starter Van
- Start date
jcrosby
Senior Member
I have Logic set to "Automatic". It never run into issues like this... I'm able to run full tracks with a TON of 3rd party plugins strapped across my mix.
The one possible difference may be that I still run Logic in Rosetta mode. Technically I probably don't even need it given that AU will bridge rosetta if/as needed, but for whatever reason I still find Logic to be less prone to unexpected crashes with Rosetta enabled. I never run into CPU issues with Rosetta enabled either... Full arrangements with orchestra, Omnisphere, Pigments, Zebra2, etc, and a ton of 3rd party plugins... All running at the same time, no overloads and very few crashes with Rosetta enabled.
Vik
Vi-k
What’s your buffer setting in Kontakt? Do you have more than one Kontakt instance per track? Do you remove (purge) the mic settings you don’t use? Not an old Kontakt version? Last Logic version?
ridgero
Senior Member
Geekbench 5
Mac Mini M2 Pro
Single Core: 1,952
Multi Core: 15,013
MacBookPro 16“ M1 Max
Single Core: 1,745
Multi Core: 12,191
Can someone close one thread? It’s annoying two use two threads with exactly the same topic.
Mac Mini M2 Pro
Single Core: 1,952
Multi Core: 15,013
MacBookPro 16“ M1 Max
Single Core: 1,745
Multi Core: 12,191
Mac14,12 - Geekbench
Benchmark results for a Mac14,12 with an Apple M2 Pro processor.
browser.geekbench.com
Can someone close one thread? It’s annoying two use two threads with exactly the same topic.
Symphonichrome
Senior Member
Deleted
Last edited:
nolotrippen
Senior Member
Look on Youtube, there's already a couple.
nolotrippen
Senior Member
Costco where I live carries Apple constantly now and will undoubtedly have the new M2s. No questions asked return policy.
Vik
Vi-k
Tro to change the buffer settings from 60k to 6k and see if that makes any difference?
nolotrippen
Senior Member
Whenever someone has a sale (looking at you 8DIO), I check what I have to see if I'm already covered. It's a fun exercise and quite often I find myself thinking, "huh, I didn't remember/know it did that."
Symphonichrome
Senior Member
Deleted
Last edited:
As I posted in another thread, please try turning off multiprocessor support in Kontakt (if that's what's causing the spikes). That got me about an extra 10% on Logic's performance meter.
I run Logic at a 64- or 128-sample buffer, by the way.
ftech
New Member
The main point of interest for me is whether the slower memory throughput of the Mini, even in M2 Pro guise, will be noticeable in real world use.
I'm still using my 2013 Pro, so either is going to be a bit more powerful anyway!
Do you have a link to where it says the memory throughput is slower? I hadn't heard that (and of course I have no idea how much of an effect it has).
And yeah, five months after getting the Mac Studio I'm still dazzled by how fast it is (coming from an upgraded 2009 Mac Pro, which is close to being as powerful as your machine).
Last edited:
Vik
Vi-k
I'm not ftech, but I found the number below through some some googling. Disclaimer: I don't know much, personally, about memory throughput – so I wouldn't know if any of my findings are wrong or misleading.
The M2 has 100GB/second memory bandwidth, the M2 Pro has a 200GB/second bandwidth.
Mac mini
Supercharged by the M2 or M2 Pro chip, Mac mini has a wide array of ports and Wi-Fi 6E for a phenomenal desktop experience.
www.apple.com
The 2009 Mac Pro doesn't even have DDR4 memory, it has DDR3 memory.
From https://community.fs.com/blog/server-ram-ddr3-vs-ddr4-vs-ddr5.html:
| Features | DDR3 | DDR4 | DDR5 |
|---|---|---|---|
| Release Date | 2007 | 2014 | 2020 |
| Clock Rate | 400–1066 MHz | 800–1600 MHz | 2400–3600 MHz |
| Voltage | 1.5 V | 1.2 V | 1.1 V |
| Transfer Rate | 800–2133 MT/s | 1600–3200 MT/s | 4800–7200 MT/s |
| Bandwidth | 6400–17066 MB/s | 12800–25600 MB/s | 38400–57600 MB/s |
One reason why the Apple Silicon memory is as fast as it is may be that it has so called 'unified memory', so this isn't only about the DDR3/DDR4/DDR5 specs.
Also:
| Max Data Rate | 1.6 Gbps | 3.2Gbps | 6.4Gbps |
Last edited:
I think the difference between M2 chips is that the higher-end one is two of the others, at least that's how the M1s work. That would account for twice everything (including the price on some machines!).
My understanding of what ftech says is that the M2 has faster memory access than the M1?
In any case, I have absolutely no idea what that means as a practical matter.
Vik
Vi-k
I'm glad I'm not alone about that.
Hansu Heya
Active Member
The funny thing is, as good as these new Apple offerings since M1 may be: the pricing of the internal SSD and (to a lesser degree) of the ram still makes me tend towards using PC-Slave for hardcore orchestral template. The reason is clearly, for most of my day to day applications the m2 Mac mini does the job and is a pretty decently priced machine. However, when you are aiming for 128gb of ram (or even more on the Mac Pro), that will also mean considerable cost for external SSDs to run all those samples on. Maybe even going for a 8tb internal SSD!
However, at this price point, you get a lower specked m2 Mac mini and could buy a pc monster on top - you will probably even have some money left. And if you use VEpro anyway, it will not even impede your workflow at all.
I just wonder, why Apple has to be THAT absurd with memory pricing?! If they wouldn’t be, I think more people would consider their top-shelf workstations …
However, at this price point, you get a lower specked m2 Mac mini and could buy a pc monster on top - you will probably even have some money left. And if you use VEpro anyway, it will not even impede your workflow at all.
I just wonder, why Apple has to be THAT absurd with memory pricing?! If they wouldn’t be, I think more people would consider their top-shelf workstations …
I think the standard M1 chip is ~66 and the standard M2 is 100. The pro and max chips from both gens are the same.I think the difference between M2 chips is that the higher-end one is two of the others, at least that's how the M1s work. That would account for twice everything (including the price on some machines!).
My understanding of what ftech says is that the M2 has faster memory access than the M1?
In any case, I have absolutely no idea what that means as a practical matter.
Vik
Vi-k
Sure, that's possible. I just read this: "While the M1 SoC has 66.67GB/s memory bandwidth, the M1 Pro has 200GB/s bandwidth and the M1 Max has a 400GB/s bandwidth". There are of course other differences between the M1 Max/M2 Max and M1 Pro/M2 Pro. They're probably both based on DDR5, which is the fastest one can get for a while (and clearly more than enough for most of us).
From:
Samsung Begins Development of DDR6 Memory: To Feature MSAP Packaging Tech With Up To 17,000 Mbps Transfer Speeds
DDR5 memory went mainstream just a few months ago but Samsung is already in its early development process of the next-gen DDR6 memory.
wccftech.com
"Samsung expects that its DDR6 design would be finalized by 2024 but commercial usage is not expected after 2025. In terms of specifications, the DDR6 memory will be up to twice as fast as existing DDR5 memory, with transfer speeds of up to 12,800 Mbps (JEDEC) & overclocked speeds blistering past the 17,000 Mbps range."
17,000 Mbps = 17 gigabyte pr. second, isn't it? Anyone here missing that? Aren't NORMI per year more important?
NORMI = Number Of Realized Musical Ideas
Justin L. Franks
Senior Member
