What's new

Ryzen 3000 Vs I9 9900K for music production

One suggestion, though. It's really really hard to discern different CPUs because you use very close shades of blue for Intels. I know this is probably because of Team Blue and Team Red

Noted, I used to do it the other way with rainbow colouring, but a user on another board told me that they felt it made more sense being able to see it by vendor platform.

Both methods have their strengths, but I must admit I agree with your comments, I was finding it tricky to differentiate the shades too.

I may change it back around on the next test round.

I noticed you ran the memories in 3200mhz
I have seen overall benchmarks for other productivity tasks and games that running memory clock at 3600mhz improves performance considerably as this also increases the clock speed of the infinity fabric bridge between the dyes on the cpu.

Will you be running any future tests to see if this could have any impact on performance for audio use?

I get asked this on every release, no it didn't in any previous testing. It helps accelerate none real-time tools like rendering or even gaming, but memory isn't the bottleneck to ASIO handling.

On Ryzen 2 I could run 2133 or 3200 and it didn't make a difference. I'll take a poke at the Ryzen 3 memory too I guess next week when I test some more mainboards, but I'm honestly expecting to see zero difference again.
 
Both methods have their strengths, but I must admit I agree with your comments, I was finding it tricky to differentiate the shades too.
I like the idea of separating by family, but I'd take more liberty with the available colors.

colors.png
 
Thanks, my collegues been telling me to stick my nose in here for years! :)

Right, testing is now live.

https://www.scanproaudio.info/2019/...00x-dawbench-tested-3-is-it-the-magic-number/

Very cool Pete! Did you happen to pay attention to the noise attribute of the x570 chipset fan?
Thanks for all your efforts. I agree with Evildragon that more colors would be welcome. Not too concerned with Team red and Team blue theme.

I was hoping to see 8700k in the Vi test but it's easy enough to compare older charts to get an idea and as far as I can tell at 512 buffer the 8700k was slightly ahead of i7 9700k in the older tests even at a lower clock speeds therefore comparing the i7 9700k to the Ryzen 7 3700x in this recent test I think it's safe to say that the 3700k has a healthy lead over the 9700k and the 8700k, plus the advantage in content creation to boot. If only it was capable of clocking to 4.9Ghz then I think it might actually beat the 9900k. Either way for the small difference the 3700x seems to be great bang for the buck compared to the 9900k.
Now I just need to watch the prime day sales and if Intel can give me a significant price drop on the 8700k then I might jump, otherwise it looks like the Ryzen 7 3700x will be the better purchase assuming I get over my fear of the chipset fan.
 
Last edited:
Very cool Pete! Did you happen to pay attention to the noise attribute of the x570 chipset fan?
Thanks for all your efforts. I agree with Evildragon that more colors would be welcome. Not too concerned with Team red and Team blue theme.

I was hoping to see 8700k in the Vi test but it's easy enough to compare older charts to get an idea and as far as I can tell at 512 buffer the 8700k was slightly ahead of i7 9700k in the older tests even at a lower clock speeds therefore comparing the i7 9700k to the Ryzen 7 3700k in this recent test I think it's safe to say that the 3700k has a healthy lead over the 9700k and the 8700k, plus the advantage in content creation to boot. If only it was capable of clocking to 4.9Ghz then I think it might actually beat the 9900k. Either way for the small difference the 3700k seems to be great bang for the buck compared to the 9900k.
Now I just need to watch the prime day sales and if Intel can give me a significant price drop on the 8700k then I might jump, otherwise it looks like the Ryzen 7 3700k will be the better purchase assuming I get over my fear of the chipset fan.

How many kontakt instances at what buffer size are you expecting to run on average?
 
How many kontakt instances at what buffer size are you expecting to run on average?

I should keep better track but at times I have over 100 tracks with a mix of Kontakt and Synths and plugins. Kontakt generally acounts for anywhere from 50% to 80% of those tracks. I'll work in buffers of 512 and sometimes more. When it comes time to record a vocal, if the song needs one, at that point I'll have usually bounced all my VSTi's but will still have a lot of plugins (waves, Izotope, etc...). Using Direct Monitoring, Reaper does a really good job of accounting for the latency even when I track a vocal at high buffers. I'm actually amazed how well Reaper handles this.
 
Noted, I used to do it the other way with rainbow colouring, but a user on another board told me that they felt it made more sense being able to see it by vendor platform.

Both methods have their strengths, but I must admit I agree with your comments, I was finding it tricky to differentiate the shades too.

I may change it back around on the next test round.



I get asked this on every release, no it didn't in any previous testing. It helps accelerate none real-time tools like rendering or even gaming, but memory isn't the bottleneck to ASIO handling.

On Ryzen 2 I could run 2133 or 3200 and it didn't make a difference. I'll take a poke at the Ryzen 3 memory too I guess next week when I test some more mainboards, but I'm honestly expecting to see zero difference again.

I am not so much after the memory boost higher frequency would bring. But the fact that it increases the speed between the dies on the cpu itself. Wouldn't the datatransfers between the dies directly on the cpu have an impact on non-partisan cpu performance?

The 3600 ram bandwidth option directly affects the die datstransfer bandwidth between the dies. That is what I am curious about if it would have an direct impact on non parallel cpu performance.
 
The ASRock X570 motherboards have Thunderbolt 3 headers that work with their own-brand Thunderbolt add-in cards. I think they are the only ones with headers.
None come with a port on the board itself from what I can tell.
 
Looking at those AsRock boards above... is the X570 limited to 64GB RAM? I know theoretically you can get 4 sticks of 32GB, but good luck finding them. Let alone at a reasonable price.


Any word yet on RTL times with the Ryzen 3000 series? If these can perform like Intel does with the Presonus Quantum, then I'll be waiting until September for that 16 core.
 
Looking at those AsRock boards above... is the X570 limited to 64GB RAM? I know theoretically you can get 4 sticks of 32GB, but good luck finding them. Let alone at a reasonable price.
The Taichi supports 128GB


Any word yet on RTL times with the Ryzen 3000 series? If these can perform like Intel does with the Presonus Quantum, then I'll be waiting until September for that 16 core.
Too early, almost nobody has one.
 
The Taichi looks like a nice board. No Thunderbolt though!
Man it's seem so difficult to get all the right specs together on one unit sometimes.

I think you just add in one of these for $100.
However I've read in a couple places that the Taichi fan is loud and whiny. Waiting to see if anyone reports a BIOS update fixing the problem. Also, whether the other (affordable) ASRock mobos have the same problem.
 
I think you just add in one of these for $100.
However I've read in a couple places that the Taichi fan is loud and whiny. Waiting to see if anyone reports a BIOS update fixing the problem. Also, whether the other (affordable) ASRock mobos have the same problem.
What kind of fan? Can it be swapped with a Be Quiet! fan?
 
The chipset fan, but someone mentioned the new BIOS fix.


From reddit
"In the BIOS you can adjust the speed to silent and it pretty resolves it.
It's been super quiet since. Either way ASROCK has some work to do on their updates."
 
Top Bottom