Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
I remember the 81xx series had a horrible jump in power usage when overclocking. Can anyone get some numbers off a Kill A Watt from stock speeds under high load and OC speeds at high load for the 83xx?
My system @ 4.8GHz is pulling ~290 watts on the Kill a Watt while running/crunching 8 threads of WCG @100% load.
I'll also look at my watt usage, when I get some watts again to use. No power since monday
Ok well I had a bit of free time today so I took a jab at the 8320. Im at 4.84Ghz currently and have been using 3dMark Vantage as a stability test, also to compare performance.
The comparison is from the best 3d Vantage CPU score I achieved with my 8120 VS 3 runs I made today. The RAM, motherboard, and PSU is the same, across all tests. SSD and waterblock have changed since the 8120 was recorded.
View attachment 117902
View attachment 117899
View attachment 117900
View attachment 117901
As you can see the 8320 handily beats the score from my 8120 @ 4.6G while running at 4.4G not a bad start. From here I bumped the multiplier 1 between the next two runs. The 4.8G run did require 2 bumps in vCore to pass, but I also reduced the LLC setting one notch so the total vCore change was +0.02v @ load.
Interestingly the graphics score scaled with the CPU speed on the 8320, it was minor, but worthwhile @ 773 points.
Total change in CPU score(8320): 1,910
Total CPU score gain from 8120: 2,621
Total change in 3dMark score(8320): 1,002
Look at the CPU score not the 3dMark score. Totally different GPU configurations for the tests, highly overclocked SLI GTX 460s, vs a single stock speed 570 thus different total scores. None the less the scaling on CPU score is still there, and as you can see continues to scale up to 4.8Ghz atleast.
Have you tried running blend on prime ? Try running it for a couple of hours and see if your OC is stable.
My AMD rig is currently down as I had to RMA my CHVF motherboard... One of the memory channels went out and the board was running in single channel mode (The CPUZ memory tab from my last posted screen shot confirms this).