CPU memory only 21% and not possible to add another amp

I know this has been something that others have also been posting about (but I can’t find the explanation). Looks like an issue that happens quite regularly after update to 2.0.0
In this example: a stock Vox AC30 Topboost on row 1 and one on row 2.
CPU usage only 21%. So I would expect enough memory left on row 1 and 2 to add another amp. But many are not possible, for instance another Vox.
Explanation? Can anyone replicate this?

When checking diagnostics Core 1 is running at about 34,5% and Core 2 (strange enough bc same amp) at 45,2. Both still way under the max. Core 3 at 5% (nothing there) and Core 4 at less than 1%…

What happens if you drag one of the amps to Lane 3 or 4 to utilize the other cores?

This was just an experiment to show that only 2 vox amps with mean about 40% of the core per row give the not enough CPU left-message.
In the presets that I make for real use I distribute as much as possible over both CPUs/ 4 cores

Are you using global EQ? If so, it is also tied to lanes 1/2, and this is probably the reason it seems different since the update

one of the first things in this experiment was to turn off the global EQ, but that didn’t make a difference.
The experiment was only to check why I get the not enough cpu message quite often since 2.0.0. When the meter is showing a low percentage, so in this case only 21%, but when I checked per row the usage per specific row (if I understand the data on diagnostics well) was imo not high enough to understand why a 3rd vox was not possible.
I will btw never do this on a real preset :slight_smile:

I’ve seen way more CPU issues since the CORS update to 2.0.0 than before. I’ve always shut off the global EQ but I think something behind the scenes is hogging the 1st and 2nd cores. I emailed support about a similar issue with adding a wah to the Johns Amp 2 preset. They didn’t give a satisfactory answer, only about not overloading cores and using processor heavy blocks… generic stuff that’s been mentioned before. The CPU monitor shows an aggregate of all the cores’ usage values, which makes sense… but what doesn’t make sense is everything you pointed above.

I hope they resolve this issue soon.I’m sure they will - seems to be software related which is good news since efficiency can always be improved. Other modelers don’t seem to have this much of an ambiguous issue with CPU usage though.