How does Cubase utilize multiple cores?

I tried it without HT enabled whilst I was setting up the PC. It’s definitely better with HT on although certainly not twice as good. If I get some time, I will do some proper testing and post numbers here.

I’ve definitely measured more benefit from clock speed than from number of cores. My setup is an orchestral template connected to three slave machines running Kontakt/PLAY/VSL via VE Pro over ethernet. I also run some instruments on the master DAW in VE Pro.

That template has been basically the same for the last 5-7 years. In that time I’ve run an i7 920 (4C/8T), i7 4930k (6C/12T), i7 6800k (6C/12T) and recently an i9 7900x (10C/20T). They all run the same template at the same latency: 96 - 128 samples (plus one more buffer for VE Pro). They were all also running overclocked between 4.0 and 4.5GHz. So the clock speeds were comparable.

So my template runs the same on 4 cores as it does on 6 cores and 10 cores that are clocked about the same. I also did tests on on those machines (except the 7900x) as slaves and found they all performed about the same.

I have also done some tests where I let them run at lower max speeds (down to 2.5 GHz) and I saw a huge performance drop. Of course there are a billion ways to measure “performance” but using number of streaming voices, number of synth patches, etc. I saw a drop of 30% - 300% depending on the metric.

So, bottom line for my setup: more than four cores has little effect. Clock speed has a huge effect.

I think the sweet spot for a DAW right now is 6 cores at 4+ GHz.


That’s funny because my Phenom II fits that description. :stuck_out_tongue:

1 Like

Fellow overclocked Phenom II x4 reporting in.
Does anyone know if overclocking the memory can help with low latency heavy load? Because I’m struggling with that.

It probably won’t make the biggest difference, but it can’t hurt. Raising the clock speed has given me the most benefits.

In games memory clock makes a lot of difference with ryzen, I’m curious about DAWs though.

I did some testing using the DAWBench ReaXComp test and the results are quite surprising.

With no HT (SMT Mode disabled in BIOS), performance is WAY better with a buffer of 32 or 64. With a buffer of 128 and above, performance is significantly better with HT enabled. With no HT, buffer size has virtually no effect on plugin instances (This is the part I find very surprising but I repeated the tests a couple of times so I’m pretty sure they’re right). Fortunately for me, I’m not concerned about low latency performance and I never run buffers as low as 64!

Funnily enough, my pervious 3 CPU’s have been an i7 920, a 4930k and a 6800k. I agree to some degree that clock speed is more important but I’ve definitely been able to run more plugins on the higher core count CPU’s -just not as many as you would expect. i.e. 12 cores certainly does not give double the plugin count than 6 cores, given the same clock speed, but it does give you MORE. That’s been my experience anyway.

I think there are so many additional variables, such as the exact plugins/routing that you use, your audio interface, your other PC components, buffer settings, any additional software you run etc etc.

and what about Mac?

2019 right around the corner… and more cores comin this year.
Either step your core game up Steinaha or get left in the dust by whoever does.

I think the marketing people att Intel are more worried than Steinaha. It is interesting that the computer has been stalling for a few years. Sort of you need to get a high end PC to get 8 core. That is what mid end phones are using. But it is not new. In 2010 there was 128 threads SPARC processors and you could have 4 of the in the same machine. They are build for servers so you can get a better performance (throughput and efficiency) if you accept some latencies. The bottle neck for multi cpu systems are usually the bus that invalidate memory caches. Memory barriers for the software people…

I built an AMD MP rig back in 2001 …actually still have it. I’m familiar with many bottlenecks…ugggh :open_mouth: That said, in 2018 not being able to have Cubase utilize an MC rig to it’s fullest is a crime. It’s almost as bad as windows users not being able to have multiple soundcards enabled in Cubase without asio4all. I mean do these damn companies talk to each other. It’s been how many years and we still have these types of “simple headaches”?..

Just frustrating…lol


Only the developers know, and they probably want to keep it secret

Multi-threaded programming is hard, really hard, and there are limited tools available to make it easy. It’s a hard problem that has been studied by programmers and computer scientists since the 70s

Some things are easy. If you have several totally independent programs running at the same time, it’s easy to efficiently use multiple cores. The key word here is independent. Even programs that appear to be independent share the disk, the screen, the network connection…etc

Processor makers hit a wall years ago when they couldn’t make a single processor run any faster, so they introduced multiple cores. The marketing department advertised them as a step forward. Programmers knew that it wasn’t that simple

If I had to guess, Cubase handles multi-core/multi-thread in an imperfect way, using best effort and a lot of secret tricks. It seems easy to imagine handing one virtual instrument in this core, one effect in another, etc. Unfortunately, the real world is a bit more messy, and unexpected timing constraints pop up all the time

I would love to read a paper by the Cubase chief software architect, clearly explaining how threads and cores are used. Methinks this will never happen

BTW, I’ve been programming since 1972, and have done a lot of work in multi-threaded realtime systems