Caporegime
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Decided to have a play with mine yesterday and I settled on stable 850mv @ 1920mhz with +500 on the memory. Tried 825mv but locks up in Time Spy straight away. 850mv passes the Time Spy test. Gives me an extra 700 points over the stock settings.
Played COD Black Ops Cold War Campaign for couple of hours gets to 66c. Noticed it was actually going to 1935mhz.
When you say 800mv at stock clocks what do you mean?I've seen people get to 800mV but at stock clocks so what you've hit is pretty good. I need to spend a bit more time on mine but even at 900mV/1905MHz I rarely see power limit throttle and temps sit at 70c even after a couple hours of gaming.
The card works in 15MHz increments so you'd expect 1920 to be correct, although maybe 1935 is the boost?
I know it's been mentioned already, but undervolting the RTX 3080 FE is free performance and an overall win-win for 1min worth of work.
Voltage/Frequency @ 900mV/1905MHz. Division 2 benchmark, went from 10,593 (118 FPS) to 10,999 (123 FPS).
When you say 800mv at stock clocks what do you mean?
You set the frequency you want at different voltages in the curve? If it's not power or temperature limited it will boost to them.
So if you set 2000 at 800mv it not really stock clocks. As you've boosted the frequency at 800mv to 2000.
I tried to undervolt but it just won't stay at the voltage, it goes lower and alters the clocks.
Ah okay I thought there was something else I was missing from the curve but essentially you just try boost higher at a lower voltage. You could call it overclocking or undervolting depending what way you want to look at the curveIt's quite an old post (from September) so the values might be relevant to when the drivers were bad; he had the clock set to 1800MHz which is what stock clocks were achieving.
Are you doing what this guys does at the start of the video? Obviously instead of his settings, you should be dragging the point at 900mV up to 1905MHz, then click apply and it'll flatten the line after it.
It should only hit that voltage and frequency under load so open HWINFO and watch the max values while playing a game.
New 3080Ti in disguise ^
What’s your Timespy graphics score with that undervolt please?
What’s your Timespy graphics score with that undervolt please?
Edit: Just gave it a go and I get 14,101. With my +105 core I get 14,488.
Maybe undervolting isn’t for me as my card runs cool anyway and my benchmark scores always seem to be better with the simple overclock.
No idea, I've never even heard of Timespy until today. I'll check it tomorrow anyway, noting that you're not convinced.What’s your Timespy graphics score with that undervolt please?
Edit: Just gave it a go and I get 14,101. With my +105 core I get 14,488.
Maybe undervolting isn’t for me as my card runs cool anyway and my benchmark scores always seem to be better with the simple overclock.
Thanks. You can get the 3DMark demo for free on Steam, which includes Timespy.No idea, I've never even heard of Timespy until today. I'll check it tomorrow anyway, noting that you're not convinced.
In my mind I'm overclocking the card but using less voltage / have lower temps, but will check that.
Coming to similar conclusions myself with my 3090.
At stock it maxes out at 62C, with a heavy OC, about 68C.
Tried undervolting and just ended up with lower scores in all benchmarks and the card ran at the same temp as at stock.
Will probably have another go at it but it was simple to work up to +160/+748 at stock PL, even with fans left on Auto.
one of the main reasons I went fe is to not support this habit.