Chilli-Man
-
Posts
120 -
Joined
-
Last visited
Content Type
Profiles
Forums
Events
Blogs
Posts posted by Chilli-Man
-
-
-
-
In for dinner and drinks
-
Anyone hit issues with OCP on volt modded cards? I had to limit my vcore to 1.33v to be able to get through a run at 1600mhz without the card turning off.
-
But then I won't be able to find anything...
-
I'm in for sure
-
LOL, let me know if you figure out how to replicate it
-
Just noticed that the rules pages mentions needing a GT 730 for stage 1, but it's not mentioned on the stage 1 page. I also think I've read elsewhere that it didn't matter what GPU was used for stage 1. I'm just using a GT 710 currently, as my 730 hasn't arrived yet.
-
Link doesn't work for me.
-
Thanks
Yes, I see that now, sorry must have missed it before.
-
Are there any restrictions on HW, like no ES CPUs or similar?
-
I've noticed the rules state each stage finishes at a different point through the month, but each stage itself says it runs for the whole month?
Also, is the z170 ln2 board allowed?
-
26th to the 3rd for me. My first computex, and to say I'm keen would be an understatement.
-
I'll take one E6300 and one E6750 thanks
-
Yes, this one. From v1b6 it's only named evc_log.txt.
Found another bug, hopefully works now https://www.dropbox.com/s/h1xmvluu483z3z6/eVc%20v1b6%20161226-3.zip?dl=0
Working nicely now
thanks -
Sorry I meant the evc_log.txt file.
This one?
-
Or try this https://www.dropbox.com/s/rub579uir5j0lct/eVc%20v1b6%20121226-2.zip?dl=0
This version should also start a lot faster and be way more conservative with memory usage.
both this version and previous seem the same. logs attached. I have to leave monitoring off, otherwise the whole program will crash and close shortly after starting the scan.
-
Thanks, found the mistake. This version should work https://www.dropbox.com/s/iua3ar8pqckyige/eVc%20v1b5%20161226.zip?dl=0
No change in monitored voltage when trying to apply an offset, but the custom command is still working.
-
Great! Try this version, hopefully monitoring is working better and added PWM frequency as well. Can you try setting Output voltage offset to 10/20/30 using the dropdown menu and record the values so I can add the mV offset instead of just numbers?
https://www.dropbox.com/s/rj6aohe4ln3d0q3/eVc%20v1b5%20161224.zip?dl=0
The monitoring dropping out is because GPUZ is reading the same data and thus "competing" for the data. If you disable monitoring in the evc software it should be fine.
Monitoring is working now, but the voltage offsets wouldn't apply. So I did it manually with the custom command, and each step was another 100mv.
-
Great! Try this version, hopefully monitoring is working better and added PWM frequency as well. Can you try setting Output voltage offset to 10/20/30 using the dropdown menu and record the values so I can add the mV offset instead of just numbers?
https://www.dropbox.com/s/rj6aohe4ln3d0q3/eVc%20v1b5%20161224.zip?dl=0
The monitoring dropping out is because GPUZ is reading the same data and thus "competing" for the data. If you disable monitoring in the evc software it should be fine.
Thanks for your time and patience with this. I'll try it out and get back to you with the values after Christmas when I get back home.
-
There's a slight chance it could break something. But I'd say the risk is relatively small and would probably try it myself.
Tried it and it worked. No lockups, and the core voltage got a small bump. It was fluctuating a little bit though, and GPUZ monitoring would drop out every so often. Losing comms?
-
Sounds like what I suspected was true. There is an alternative register you could try but I'm not a 100% sure what it does. If you don't want to risk your card you should not try it.
If you're up to the task however, try this using "Send custom command" from the Menu:
Enter Address 20 Register E6 Data 10 ("R/W Word" unchecked), press Write and check if the output voltage changed.
What happens if it's wrong? And would you try it on your own card? Happy to take the gamble if you think there's good odds it'll work.
-
Everything seems correct. I suspect that the GPU doesn't like that it's not seeing the VRM controller over SVI anymore (this method is changing the control to SMBus instead). Can you measure the output voltage and what happens to it when you press apply?
Tried applying a voltage and I didn't get any change in the output voltage when measuring it with a multimeter. PC just locked up and the GPU voltages stayed the same.
-
I'll take one of each if there's any left thanks
RULE - Core i9 7900X @ 5900MHz - 4506 marks XTU
in Result Discussions
Posted
Awesome!