Sign in with
Sign up | Sign in
Your question
Solved

AMD 955 BE VCore oddity

Last response: in Overclocking
Share
August 7, 2011 4:39:18 PM

Hello All!

I've recently built my first computer (heavily influenced by the SBM budget models), specs here:

CPU: AMD Phenom II 955 BE (3.2 GHz Quad-Core)
Motherboard: ASRock M3A770DE
RAM: G.Skill DDR3-1600Mhz 4GB
GPU: Sapphire Radeon HD 6850
HDD: 500GB WD Caviar Blue
CPU Cooler: CoolerMaster Hyper 212+
Case: Antec 300

It's been running stably for days now at stock clocks, so I just recently upped the multiplier to 16.5x (3300 MHz clock). It is running stably, with its low temperature being about 33-35° C. However, the Prime95 Blend Test pushes the temperatures to 43-45°C, which seems rather high for such a small clock bump. When I checked CPU-Z's VCore reading, it listed VCore as 1.45V, which seems unnecessarily high.

The only steps I took to overclock were to enter the BIOS->Advanced->CPU Configuration and changed the "CPU Multiplier" from Auto to Manual and stepped it up to 16.5x. Cool'n'Quiet was automatically disabled once CPU Multiplier was set to Manual, so I'm wondering why I have such a high VCore. Any solutions? Should I undervolt this or change a setting somewhere?

*On a side note, I am using Arctic Silver 5 (combo pricing on Newegg made it worth it), and I'm reading that I'll have higher-than normal temperatures until it cures.

Thanks!

More about : amd 955 vcore oddity

Best solution

August 8, 2011 2:28:50 AM

PurpleHayes said:
it listed VCore as 1.45V, which seems unnecessarily high

I was using the M3A770DE last year and I don't remember having a "VCore" option to change, if I remember correctly it was just a direct "CPU Voltage" option along with CPU-NB, HT, and others.

For 3.7GHz I have my 955 @ 1.4v but anything under 3.5GHz I run completely stable @ 1.25v.
My CPU @ 3.7GHZ idles around 35-37°C, with Prime95 running stays between 48-52°C.
@ 3.2GHz with 1.25v after an hour or so of Prime95 my temps got up to 48°C so you're not doing too bad temperature wise even if your voltage is way too high.

PurpleHayes said:
Should I undervolt this or change a setting somewhere?

I would suggest setting your CPU voltage at 1.25v and see how high you can get the multiplier before unstable then do small .025v bumps and repeat. The lower the setting the less heat you're going to be generating also.

If you don't like the temps you're getting, and you haven't already, you should fill the rest of the fan slots in your Antec 300. The 300 Illusion I have keeps everything pretty cool with it's five fans (3x 120mm intake, 140mm & 120mm outtake).
Share
August 10, 2011 4:20:55 AM

Awesome, thats everything I needed and more :)  Many thanks, pushed it to 3.6 and doing a 24-hour Prime95 :D 
m
0
l
Related resources
August 10, 2011 2:38:07 PM

PurpleHayes said:
Awesome, thats everything I needed and more :)  Many thanks, pushed it to 3.6 and doing a 24-hour Prime95 :D 

cool, what voltage did you settle at for the 3.6GHz?

if you're comfortable with overclocking, and haven't yet, you might wanna try pushing the NB up to 2600 or 2800MHz. it can boost the speed between your memory and CPU and other components giving you higher rates of data transfer. My NB is at 2600MHz with the CPU-NB voltage set at 1.265v and all my benchmarks and FPS in games went up noticably. if you have the C3* revision of the 955BE you can probably get your NB to 2600MHz with even less than 1.265v.

*C3 model number: HDZ955FBGMBOX.
my C2 version (HDZ955FBGIBOX) needs quite a bit more voltage and still won't run at quite as high speeds as the C3.

come to my site and post your overclocks:
Overclockers Reference
m
0
l
August 17, 2011 4:13:14 PM

Hmm...CPUID says I have revision RB-C3 stepping (so C3). I've got it at 1.275 just because I want to ensure stability, my next goal is to find some way to establish custom P-States so I can emulate Cool'n'Quiet (prolong processor life?).

I think after that I'll try pushing the NB, but here's a follow-up question: I know with the multiplier I'll get glitchy/BSOD screens if I push it too far, but doesn't changing the NB necessitate clearing the CMOS if you go too far?
m
0
l
August 17, 2011 6:23:52 PM

PurpleHayes said:
my next goal is to find some way to establish custom P-States so I can emulate Cool'n'Quiet (prolong processor life?).

download K10STAT. it's an awesome program you can use to create your own custom "cool & quiet" profiles that run from inside Windows. I've been using it for years now and never had a problem. a new version just came out, you can download it here:
Overclockers Reference.com-Help Section-Downloads-K10STAT v1.54
Also you can be one of the first to post your CPU specs at our new site on the AMD Chipset Page. Please Contribute.
Overclockers Reference-AMD Chipsets


PurpleHayes said:
I think after that I'll try pushing the NB, but here's a follow-up question: I know with the multiplier I'll get glitchy/BSOD screens if I push it too far, but doesn't changing the NB necessitate clearing the CMOS if you go too far?

Your Motherboard should have an automatic setting that if it doesn't boot correctly after 2-3 tries it will reset the BIOS on it's own. But you can disable this function if you wish. After the first failed boot you can just enter the BIOS again before it loads and turn your NB multiplier settings back down or it's CPU-NB voltage up. If it's pushed too far for your current voltage usually Windows will still load up and crash right away or have the same instability it would have from the CPU multiplier being too far.
m
0
l
August 20, 2011 11:09:16 PM

Best answer selected by PurpleHayes.
m
0
l
!