• Competitor rules

    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.

Underclock / Undervolting an 8350

Associate
Joined
17 Apr 2010
Posts
461
Hi,

Has anyone successfully underclocked / undervolted an 8350 and what power savings were you able to achieve?

If you could share your settings to achieve your undervolt if this is what you did that would be appreciated.

Thanks,
Chris.
 
Soldato
Joined
12 Sep 2012
Posts
11,696
Location
Surrey
This is an interesting thread if it takes off. Not because it makes the CPU greener but because it will shed some light to what the CPU is like at the other end of the spectrum than we normally talk about on here.

Inb4 a fanboy makes this discussion about alternative power saving CPUs; very few of us actually have our CPU running under load constantly, so posting a review comparing power usage of Haswell to Piledriver is irrelevant. I would quite like to see some 8350 screens, so lets not veer this thread away!

My current two 8350's are not amazing chips in terms of clocking, so i dont think i can reduce the voltage too much before it decides that i need to knock the multiplier down a fair bit but will have a go later. On the other hand, a 8150 i was sent aaaaaaages ago was a stunning chip though and i even pumped the clock UP rather than drop it when i lowered my voltage to 1.25V.

525810_10151276103918072_1798063239_n.jpg

I will try this later on my 8350's but i am pretty sure i cant even get it to boot at stock clock when my voltage is that low, LOL
 
Soldato
Joined
10 Oct 2012
Posts
4,415
Location
Denmark
My current two 8350's are not amazing chips in terms of clocking, so i dont think i can reduce the voltage too much before it decides that i need to knock the multiplier down a fair bit but will have a go later. On the other hand, a 8150 i was sent aaaaaaages ago was a stunning chip though and i even pumped the clock UP rather than drop it when i lowered my voltage to 1.25V.

525810_10151276103918072_1798063239_n.jpg

I will try this later on my 8350's but i am pretty sure i cant even get it to boot at stock clock when my voltage is that low, LOL

Jeezes..wish my PD could do that and i even think i might have a good chip already. Think the lowest i got the volts so fare on my 8350 at stock would be around 1.33-1.32v but then again i have barely been running it at stock so this could be a fun thing to try out.
 
Soldato
Joined
12 Sep 2012
Posts
11,696
Location
Surrey
Jeezes..wish my PD could do that and i even think i might have a good chip already. Think the lowest i got the volts so fare on my 8350 at stock would be around 1.33-1.32v but then again i have barely been running it at stock so this could be a fun thing to try out.

Haha, i know i got lucky with that one.

I have 2 8350's on much better motherboards and they cant do what the 8150 is doing. The annoying thing is that the 8350's both have watercooling that i threw money at, to really stretch the clock, and they don't go much higher but need to be overvolted to 1.47-1.5 to hit 5.1GHz. The 8150 is on half decent air cooling and unfortunately i need to go all the way to 1.42V to go past the screenshot clock. One day i will have the motivation to swap the 8350 out of my Formula-Z board for the 8150 for some tests. My stupid obsession with making the tubing routes as neat as possible between RAM, CPU and VRM block makes it impossible to take out the CPU without draining the loop and redoing the tubing.

Edit:

This motherboard is the one i screened with. The only reason why i tried undervolting in the first place is because the stock voltage was killing the super crap VRMs. Got a feeling that if i tried this with an equivalent price Asrock board first, i would have an MicroATX Paperweight with medium-welldone MOFSETs
 
Last edited:
Soldato
Joined
12 Sep 2012
Posts
11,696
Location
Surrey
I don't think it will throttle the voltage and not the multiplier with it. I had to run the lower voltage because the motherboard VRMs were dying at even stock voltage under load. You should see my DIY VRM cooling attempt, it's like a birds nest of coathanger wire cradling a noisy high speed 80mm fan directed at the VRMs. Sure the screen looks impressive but if you saw my test bench, it looks like junk glued to other junk. Stability wise, it lasts about half hour to hour+ before workers shut down randomly in prime, it didnt struggle with any other benches or stresses but i will say that the benchmark score was a little lower on the 4.8 clock than a 4.5 clock on a slightly higher voltage and a mixture of both multiplier and bus speed clock. So in terms of numbers on paper, yeah it is cool but didnt perform any better.
 
Soldato
Joined
10 Oct 2012
Posts
4,415
Location
Denmark
i see, well we all know how prime95 abselutely rapes the AMD FX line, so i wonder how it would do with a stress test that actually stressed it properly. Never the less this is what being a geek is all about. Love it. Junk clued to junk is describes my first pc(scraps found from the scrap yard) and screwed together with a potato peeler(for real, im serious here :p)
 
Associate
OP
Joined
17 Apr 2010
Posts
461
Testing so far, running all 8 cores at 3.5 ghz (fx-8320 speeds with turbo boost disabled) and vcore down to 1.2, still testing but prime stable, going to push it lower soon, upped LLC to combat vdroop.







Going to get a Kill-A-Watt Electricity monitor today.
 
Last edited:
Back
Top Bottom