I have an msi motherboard, i think LLC is set to 4.. i had successfully gone through stress tests p95 blend with fixed bios setting, i think it was for nearly 16 hours, vcore under load of around 1.308v (bios was around 1.32 i think), this at 50x.
Then i tried adaptive, i tried just regular adaptive, put the volts in with no offset, volts during load would hit 1.37 or 1.38, same deal with adaptive + offset set to auto (tried stricter LLC of 2 as well), then i tried manual negative values and got the load values close to the 1.308 above.
For weeks i had vcore in bios at 1.350 and negative -0.05 adaptive. Suddenly during gaming the pc reset itself. I dont think i had the bsod setting correct because there was none recorded.
This same configuration was yielding i think around 0.675 min and during occt adaptive, it was around 1.313. I'm guessing a "low" volt moment caused it, not the load one.
So at this point im debating what to do, raise bios volts to 1.36 and keep neg 0.05, or lower to 1.345 and go with 0.04 or similar? Then do at least 2 hours small fft with occt to see if stable? Followed by maybe another prime95 blend? Or maybe blend is better since there will be periods of lower volts. . Or maybe cant stress test, perhaps this means the load vcore i thought was stable (1.308v at fixed) just wasnt.
Anyone have any suggestions? I believe i need to find the vmin and stay above it, but i dont completely understand the terminology at this point. Sure I could put back the fixed that worked and leave it (having only the clock speed downclock), but i was trying to go for gold so to speak and get some energy savings.
Thanks in advance
Then i tried adaptive, i tried just regular adaptive, put the volts in with no offset, volts during load would hit 1.37 or 1.38, same deal with adaptive + offset set to auto (tried stricter LLC of 2 as well), then i tried manual negative values and got the load values close to the 1.308 above.
For weeks i had vcore in bios at 1.350 and negative -0.05 adaptive. Suddenly during gaming the pc reset itself. I dont think i had the bsod setting correct because there was none recorded.
This same configuration was yielding i think around 0.675 min and during occt adaptive, it was around 1.313. I'm guessing a "low" volt moment caused it, not the load one.
So at this point im debating what to do, raise bios volts to 1.36 and keep neg 0.05, or lower to 1.345 and go with 0.04 or similar? Then do at least 2 hours small fft with occt to see if stable? Followed by maybe another prime95 blend? Or maybe blend is better since there will be periods of lower volts. . Or maybe cant stress test, perhaps this means the load vcore i thought was stable (1.308v at fixed) just wasnt.
Anyone have any suggestions? I believe i need to find the vmin and stay above it, but i dont completely understand the terminology at this point. Sure I could put back the fixed that worked and leave it (having only the clock speed downclock), but i was trying to go for gold so to speak and get some energy savings.
Thanks in advance