Open LuisB79 opened 1 year ago
Hi Sp00n,
First of all, thanks for creating this awesome tool that makes the job much easier for many overclocking/overclockers out there. I was fascinated by the idea and how you developed the script along with loggers, testing tools etc.
I just came across a tool called SMUDebugTool which is hosted in Github in the below URL:
https://github.com/irusanov/SMUDebugTool
When I was going through a post in www.overclock.net I found this where they have included a tab for PBO, which could be used for tuning the PBO Curver Optimizer for Ryzen 7950X/7950X3D. Kindly include it (or) suggest it in your read me as it can quickly make changes to the Curve Optimizer without having to boot to BIOS just like PBO2Tuner (since PBO2Tuner doesn't work on 16 cores and with Raphael).
Of course, you can use your testing before suggesting it, it will be a complete package for any Ryzen Processors.
i have some suggestions for features /improvements 1.-Could it be possible to integrate hwinfo data in log files? 2.-Could it be possible to have a setting where instead of a time limit, the core would end its testing based on the iteration completion on y-cruncher? or have the 2 work together? sometimes y-cruncher doesn't respect its 60second time limit set. 3.-Could it be possible to have a script for automatically adding -1 or +1 with ryzenadj in CO value depending on the stability on a y-cruncher run with a certain set of iterations, for fastish testing, or overnight testing?, where if it fails and the system reboots, the script would pick on based on the last log file, probably some stability charts could be made with that.
1) Unlikely, this is a bit out of scope of the script. 2) The 0.9.5.0 alpha versions already have an "auto" setting for y-Cruncher, where the script waits until all the tests have been completed for a core. 3) There's an edit floating around in the overclock.net thread, that uses PBO Tuner to automatically set the CO value +1 when an error occurs. It's on my todo list to be added eventually, but I can't give an ETA.
@LuisB79 Point 3 is now included in the 0.10.0.0 alpha, check it out.
I did notice that you mentioned ryzenadj, which seems to be a mobile Ryzen only tool. At least it didn't want to work together with my 5900X ("Only Ryzen Mobile Series are supported"). So if you're using a mobile chip, I could actually use your feedback if the current implementation for Automatic Test Mode also works for mobile CPUs, or if I need to add the RyzenAdj tool as well to support these (or make PJVol adjust his tool 😄).
oh yeah i used a 5900x, but now i have dragonrange 7945hx which has 16 cores, i could test it out none the less
i have some suggestions for features /improvements 1.-Could it be possible to integrate hwinfo data in log files? 2.-Could it be possible to have a setting where instead of a time limit, the core would end its testing based on the iteration completion on y-cruncher? or have the 2 work together? sometimes y-cruncher doesn't respect its 60second time limit set. 3.-Could it be possible to have a script for automatically adding -1 or +1 with ryzenadj in CO value depending on the stability on a y-cruncher run with a certain set of iterations, for fastish testing, or overnight testing?, where if it fails and the system reboots, the script would pick on based on the last log file, probably some stability charts could be made with that.