Hello,
I bought a EVGA GeForce GTX 1080 FTW GAMING ACX 3.0 and have it overclocked. +80 offset on the core clock, +350 offset on the memory clock, so I'm running at 2076 MHz on the core clock & 5355 MHz on the memory clock.
I'm finding it really frustrating to overclock because this card acts like it has a mind of its own and runs its core clocks at different ranges than I've specified.
When overclocking ins MSI Afterburner, I run Unigine Heaven Benchmark on a loop in the background to keep my GPU under load. When I overclock, I've been going up in small increments until my video driver crashes, then back down to the last stable increment. I then overclocked my memory prior to the point of seeing artifacts, then revert back to the last increment without artifacts.
However, many times I've seen my core clocks start at over 2114 MHz in Unigine Heaven and then slowly settle down to 2076. If I tried to overclock to 2114 MHz my video driver will crash so I don't know why my card is ramping to 2114 MHz on its own automatically and does so with stability, yet I can't get there manually.
When nothing is running, GPU-Z's PerCap Reason states:
Util = Utilization. Indicating perf is limited by GPU utilization.
Then once I start Unigine Heaven there's no reason given and my GPU core clocks are at 2114.
For the next two minutes of running, the GPU core clock start to drop incrementally down to a resting 2076. During this time GPU-Z PerfCap Reason states:
vRel = Reliability. Indicating perf is limited by reliability voltage.
I have my power flimit maxed to 120%, my voltage unlocked and maxed on on MSI Afterburner, though I don't know if the Voltage/Frequency Curve found by (Ctrl+F) is overriding it..
I think GPU boost is a culprit here and it's making it a pain to overclock because it's throwing my clocks around at numbers that are stable or unstable.
I bought a EVGA GeForce GTX 1080 FTW GAMING ACX 3.0 and have it overclocked. +80 offset on the core clock, +350 offset on the memory clock, so I'm running at 2076 MHz on the core clock & 5355 MHz on the memory clock.
I'm finding it really frustrating to overclock because this card acts like it has a mind of its own and runs its core clocks at different ranges than I've specified.
When overclocking ins MSI Afterburner, I run Unigine Heaven Benchmark on a loop in the background to keep my GPU under load. When I overclock, I've been going up in small increments until my video driver crashes, then back down to the last stable increment. I then overclocked my memory prior to the point of seeing artifacts, then revert back to the last increment without artifacts.
However, many times I've seen my core clocks start at over 2114 MHz in Unigine Heaven and then slowly settle down to 2076. If I tried to overclock to 2114 MHz my video driver will crash so I don't know why my card is ramping to 2114 MHz on its own automatically and does so with stability, yet I can't get there manually.
When nothing is running, GPU-Z's PerCap Reason states:
Util = Utilization. Indicating perf is limited by GPU utilization.
Then once I start Unigine Heaven there's no reason given and my GPU core clocks are at 2114.
For the next two minutes of running, the GPU core clock start to drop incrementally down to a resting 2076. During this time GPU-Z PerfCap Reason states:
vRel = Reliability. Indicating perf is limited by reliability voltage.
I have my power flimit maxed to 120%, my voltage unlocked and maxed on on MSI Afterburner, though I don't know if the Voltage/Frequency Curve found by (Ctrl+F) is overriding it..
I think GPU boost is a culprit here and it's making it a pain to overclock because it's throwing my clocks around at numbers that are stable or unstable.
Facebook
Twitter
Instagram