- Jun 12, 2014
- 2,125
- 24
- 12,465
So... I overclocked my FX-8320 to 4.4 GHz with a hyper evo 212 as discussed in this thread. As of right now, I was able to push the CPU to a stable 4.4 GHz with a voltage of 1.38v. I'm hitting a thermal wall now since my CPU is reaching close to 62 degrees, and at the same voltage, 4.5 GHz gives errors in prime. A higher voltage makes my CPU too hot when running Prime95 with all 8 cores. So with my current cooling setup I'm limited to 4.4 GHz.
There is still room to increase the voltage to increase stability, since the maximum allowed (although a bit high) is generally 1.5v. Default boost voltage is even 1.42. The problem is my temperatures. I did the best I could with air cooling as of now. I've tried looking up on the internet if other people have enabled boost after overclocking, but, it seems that people simply leave it disabled, which I think is a shame since I see a possibility for more performance.. I wonder if there's a specific reason no one does it...
In any case... Since turbo boost generally only works for a short time, and the maximum boost applies only on one or two cores simultaneously, I want to determine which turbo boost settings are feasible for my system. I plan to do the following;
1) I'll increase the voltage one step to start off.
2) I'll increase the frequency to 4.5GHz to start off.
3) Rather than testing all 8 cores at the same time with Prime95, I want to test 4 cores first, then the other 4 cores, to see how it goes with my temperatures.
4) In case the system still gets too hot with the method above, I'll have to try with 2 cores instead. I personally don't see it as a problem since the maximum boost is only allowed on 2 cores simultaneously, and is generally only enabled in case of a heavy single threaded application. So I'll be testing the first two, then the next two, and so on until I've covered all the cores.
5) In case of failure, I'll increase the voltage one step as long as there's headroom for it, and repeat until all the cores are stable.
6) I'll be increasing the voltage and frequency as things progress. I'll push the clock as far as I can get it within the maximum allowable voltage, and each core must past. Obviously keeping my temperatures in check.
7) I'll change my boost settings in BIOS with the maximum stable clock I found at the specific voltage that it was stable.
I want to know what you guys think about this, before I try it.
- Is there something I'm missing?
- Why hasn't anyone done/tried this before?
- Is there any reasonable and important reason why I should NOT do this?
Relevant System info:
CPU: FX-8320
CPU Cooler: Cooler Master Hyper 212 EVO
Cooler on backplate: 80 mm fan from the stock cooler that came with the CPU
Motherboard: Asus M5A99FX Pro R2.0
Memory: 4x 2GB Kingston ValueRAM 1333 CL-9 1.5V @ 1600 CL-9 1.55V
Case: Fractal Design Arc Midi R2 (Stock 140 mm intake fan in front, intake fan moved to front bottom instead of front middle. 140 mm stock outtake fan in back and on top. Additional Cooler Master 120 mm fan intake installed in bottom of case.
PSU: EVGA SuperNOVA 750G2
There is still room to increase the voltage to increase stability, since the maximum allowed (although a bit high) is generally 1.5v. Default boost voltage is even 1.42. The problem is my temperatures. I did the best I could with air cooling as of now. I've tried looking up on the internet if other people have enabled boost after overclocking, but, it seems that people simply leave it disabled, which I think is a shame since I see a possibility for more performance.. I wonder if there's a specific reason no one does it...
In any case... Since turbo boost generally only works for a short time, and the maximum boost applies only on one or two cores simultaneously, I want to determine which turbo boost settings are feasible for my system. I plan to do the following;
1) I'll increase the voltage one step to start off.
2) I'll increase the frequency to 4.5GHz to start off.
3) Rather than testing all 8 cores at the same time with Prime95, I want to test 4 cores first, then the other 4 cores, to see how it goes with my temperatures.
4) In case the system still gets too hot with the method above, I'll have to try with 2 cores instead. I personally don't see it as a problem since the maximum boost is only allowed on 2 cores simultaneously, and is generally only enabled in case of a heavy single threaded application. So I'll be testing the first two, then the next two, and so on until I've covered all the cores.
5) In case of failure, I'll increase the voltage one step as long as there's headroom for it, and repeat until all the cores are stable.
6) I'll be increasing the voltage and frequency as things progress. I'll push the clock as far as I can get it within the maximum allowable voltage, and each core must past. Obviously keeping my temperatures in check.
7) I'll change my boost settings in BIOS with the maximum stable clock I found at the specific voltage that it was stable.
I want to know what you guys think about this, before I try it.
- Is there something I'm missing?
- Why hasn't anyone done/tried this before?
- Is there any reasonable and important reason why I should NOT do this?
Relevant System info:
CPU: FX-8320
CPU Cooler: Cooler Master Hyper 212 EVO
Cooler on backplate: 80 mm fan from the stock cooler that came with the CPU
Motherboard: Asus M5A99FX Pro R2.0
Memory: 4x 2GB Kingston ValueRAM 1333 CL-9 1.5V @ 1600 CL-9 1.55V
Case: Fractal Design Arc Midi R2 (Stock 140 mm intake fan in front, intake fan moved to front bottom instead of front middle. 140 mm stock outtake fan in back and on top. Additional Cooler Master 120 mm fan intake installed in bottom of case.
PSU: EVGA SuperNOVA 750G2
Thanks guys.
Facebook
Twitter
Instagram