'Quick initial test yCruncher' Query

[OZJZERO]

I have a Ryzen 9800X3D.

How long is this test supposed to take? Is it an indefinitely looping test or does it expire after a certain time limit or number of iterations?

So far I've run this test and it's still running for 2 hours.

[sp00n]

Yeah, there is no real time limit for these tests, so it'll basically run indefinitely.
Insofar the "quick" in the config file's name might be a bit misleading, but the whole philosophy of CoreCycler is based on not providing some arbitrary amount of testing time and then label that as stable or not, instead the user can decide for themself if it's good enough or not.

So it's not holding your hand as much, but allows for much greater freedom.

[OZJZERO]

Thanks for the reply.

What makes this 'quick initial test' different to the Breadbit YCruncher script? The wording of the file made me assume this was a short or time limited test to 'screen' for any errors.

[sp00n]

The BreadPit config files are a bit more specific, they were developed with and for a 5800X3D and may not be as effective for other processors, which is why I added a more general y-cruncher config as well.

The "quick" in the name points towards the experience that y-cruncher with Kagari seems to find obvious errors pretty quickly, but I do see that it can cause confusion.

[OZJZERO]

Interestingly I've run every single stress test but the only one which produces errors is the 'Low load scenario Prime 95'. I thought I was lucky to have -30 all core undervolt but nope. I've had to dial back to -25 on some of my cores.

Would this suggest that I am more likely to experience crashes / BSOD on casual browsing?

[sp00n]

It's hard to tell. Normally the light load / huge FFTs in Prime95 causes the boost clocks to go the highest, and this could find instabilities that couldn't be seen before for harder work loads, because the clock speeds never could get that high. This and the load changes to and from idle are normally the main causes for crashes during browsing and other low load usage scenarios.

However the 9800X3D should be able to reach its highest single core boost clock even during multi core load right out of the box, so higher frequencies shouldn't be the main issue here. To be honest, I don't really know, and there hasn't been that much feedback for the 9800X3D yet (and I don't follow the respective threads actively either).