i5 11600k vs i7 10700k vs i9 10850k

[Eagle Eyed Jerry]

Hello all,

I'm looking at building a new PC for music production and I'm struggling to decide whether I should go for an 11th gen i5 or a 10th gen i7/i9.

The i5 supports PCIe4 (so can make use of a 980 Pro SSD vs the 970 EVO Plus for the 10th gen Intel's), has better inegrated graphics and has higher single core clock speeds, whereas the i7/i9 has more cores. I'm not sure what is more important!

Should I be prioritising single core speeds or number of cores?

My typical workflow is to record guitar/bass and use NeuralDSP amp sims then use Kontact drum samples for guide tracks before re-recording with live drums. I often add other VST instruments like synths/pianos too. 

I've attached some example builds bellow, if anyone has any thoughts on them, please let me know!

Thanks!

i5 11600k: https://uk.pcpartpicker.com/list/fQyHt8 ~ £870
i7 10700k: https://uk.pcpartpicker.com/list/Th9wfP ~ £880
i9 10850k: https://uk.pcpartpicker.com/list/N2kHt8 ~ £940




 

Edited April 6, 2021 by James Holmes

[craigb]

Jim Roseberry (member and owner of Studio Cat / Purrrfect Audio ) has written about this several times here.  I'll see if I can find one of the posts (and, maybe, Jim can jump in for an answer too!).

[craigb]

Found a few that may be of interest:

 

[Jim Roseberry]

For working at ultra low latency settings, clock-speed is the single most important factor.

ie:  Some audio interfaces like the Antelope Orion Studio Synergy Core will allow you to run at 96k using a 32-sample ASIO buffer size... resulting in 1ms total round-trip latency.  Running amp-sim plugins at these settings isn't something that lends itself to being heavily multi-threaded (spread across multiple cores).

More cores is certainly beneficial (especially at higher buffer sizes), but not at the expense of significant clock-speed.

In a perfect scenario, you want highest clock-speed... AND the most cores you can get.

 

Tested the 11900k recently.

It's a performance improvement (vs the 10900k) in most areas... but not all.

There were some changes in the CPU architecture (for Rocket Lake) that are a bit more latent.

Working at larger buffer sizes, you'd not notice.

If you're trying to run Neural DSP plugins, Helix Native, etc... at 96k using a 32-samples ASIO buffer size (or smaller), you'll hear glitches.

That's the one area Rocket Lake is a step backward (ultra low latency audio).

 

For the performance/cost, it's still hard to beat the 10900k.

[Amicus717]

2 hours ago, Jim Roseberry said:

For working at ultra low latency settings, clock-speed is the single most important factor.

ie:  Some audio interfaces like the Antelope Orion Studio Synergy Core will allow you to run at 96k using a 32-sample ASIO buffer size... resulting in 1ms total round-trip latency.  Running amp-sim plugins at these settings isn't something that lends itself to being heavily multi-threaded (spread across multiple cores).

More cores is certainly beneficial (especially at higher buffer sizes), but not at the expense of significant clock-speed.

In a perfect scenario, you want highest clock-speed... AND the most cores you can get.

 

Tested the 11900k recently.

It's a performance improvement (vs the 10900k) in most areas... but not all.

There were some changes in the CPU architecture (for Rocket Lake) that are a bit more latent.

Working at larger buffer sizes, you'd not notice.

If you're trying to run Neural DSP plugins, Helix Native, etc... at 96k using a 32-samples ASIO buffer size (or smaller), you'll hear glitches.

That's the one area Rocket Lake is a step backward (ultra low latency audio).

 

For the performance/cost, it's still hard to beat the 10900k.

This is great info,  thanks Jim. I was wondering about this myself, as I’m shopping for a new build.