Assistance with boosting 8320e beyond 1.32V
Assistance with boosting 8320e beyond 1.32V
I've been working on this for several days yet I still can't solve it.
Specs: FX 8320e
Corsair H80i
gigabyte 970 gaming sli
corsair vengeance 16Gb(2x8GB) 2400Mhz running at 1866Mhz
EVGA 1070 FTW
EVGA 600W B1 80+ bronze
OS: Windows 10 home creator's update
During stress testing if I have vcore set to anything over 1.32V once the test has been running for about 20 minutes the vcore will drop to about 0.98V and the CPU will downclock to 1.4GHz, it will then jump back up to the proper speed and voltage. The longer the test runs, the shorter the interval between the drops. It shouldn't be an issue with temps as the CPU never goes past 50C and all the motherboard temps never go past 55C.
I've turned off all unnecessary CPU features in the BIOS (turbo, cool&quiet, etc.) and enabled HPC, I've tried adjusting the LLC settings, but the drops still occur no matter what the LLC is set to. I've set windows power settings to high performance and disabled fast boot.
I've achieved a fairly stable overclock at 4.0GHz and 1.32V but would like to increase it further.
Is this solely a hardware problem with my motherboard or PSU, or am I making a mistake?
Edit: Forgot to note that I acquired this CPU second hand from a friend who was running it at 4.6GHz 1.4V without any problems.
You can increase vcore up to 1.55 on visheras while keeping the cooling stable. If you're seeing Vdrop, check if the motherboard supports an LLC and turn it on.
The No 970 chipset motherboard offers sufficient VRM headroom for significant overclocking of an FX octacore, though doing so may not always be successful. Raising LLC settings only exacerbates the issue. It seems you've reached the thermal, voltage, or current limits of your motherboard's VRM, which can cause throttling. Consider upgrading to a more powerful board such as the 990FX or ASRock Fatal1ty, or switch to Ryzen processors. Your friend likely has a better power phase design and VRM setup. If you shared my FX8320 that reached 5.2ghz with phase change cooling (it failed at 5.1ghz), does that suggest it could also run at 5.2ghz on less capable hardware?
Amtsseung :
No 970 chipset motherboard offers sufficient VRM headroom to push an FX octacore overclock significantly, if possible. Raising LLC settings only exacerbates the issue. It seems you've reached the thermal, voltage, or current limits of your VRM, which is causing throttling. You might need a more powerful board—such as the ASRock Fatal1ty 990FX Professional or ASUS Sabertooth 990FX—or consider switching to Ryzen processors. Your friend likely has a better power phase design and VRM setup. If I had my FX8320 overclocked to 5.2ghz using phase change cooling (it actually failed at 5.1ghz), does that imply you could achieve the same with lower-quality hardware? Ohh rly? You own a 970 gaming card that performs better than any 12-phase crosshair kitty or Holzerock.
And it handles quads, six, and octa cores without issues at higher TDPs.
I think you may have reached the thermal and/or voltage and/or amperage limits of your motherboard VRM, which is known as VRM throttling. You might want to upgrade to a more powerful board, such as the 990FX models from ASRock or ASUS, or consider switching to a Ryzen processor. Your friend likely has a better power phase design and VRMs. The board they were using had an 8+2 power phase, similar to yours, so it could just be an issue with your own VRMs. Unfortunately, it's out of warranty.
Time to start saving for a Ryzen, at least. At least I'm running the CPU at 4GHz without any problems—it's not bad for a CPU I got for $20.
Did you actually kill those 2 8350s? How? I didn’t do any chip OCing, I only fried a VRM once on an old UD3 early revision CPU—never a CPU. Not even the cheap used 6300 I bought on eBay. I ran it at 4.6 and even went overvoltage past 1.55, but nothing happened. It was stable all the time.
4 phase?!?!?!!?
I owned a rare Sapphire 790x board from Europe once. After trying to overclock a PII 940 black, it wouldn’t move past 100 MHz, but it handled 4 phase without any issues. Probably because I didn’t push it too hard—it didn’t even need a heatsink on the VRM.
MylegsareOK :
I was quite clueless and unlucky, honestly. Just a bit more experience would help. When I was younger, I accidentally zapped two Tbirds with static... it happened when I tried to do a pencil trick on top of a pack of cigarettes. The cellophane paper had so much static! I didn’t realize it at the time.
😉