• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Really puzzled by new X4 955 BE

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Frozenwings

Member
Joined
Jul 28, 2005
I just got all my new parts in a couple days ago and I've been toying with OCing. I would like to consider myself a veteran OCer, I've been doing it for quite a while, although with all this CPU, CPU-NB, HTT, NB stuff it's been quite the adventure.

Here's what we're working with:

AMD Phenom II X4 955BE
Sunbeam Core Contact Freezer
MSI 790FX-GD70
Nvidia 275GTX
2x PC1600 Patriot RAM (7-7-7-20, 1.9v)
Corsair 750w PSU

Anyways, the problem I'm having is that I just can't seem to get good temps or OCs no matter what I do. I have yet to get a BSOD or a restart, but I can not for the life of me get this thing Prime stable running blends at just about any sort of OC. I had to crank up the CPU-NB voltage to 1.25v to get it stable at stock settings to run Prime for over 10 hours without errors.

At stock settings I'm seeing about 35C idle and 50C load, which appears kind of hot for the heatsink I have. I've read that the phenoms get a little funky when you get over the 50-55C mark as far as overclocking goes. So that there could be the issue. In the same room and same computer case with my past 2 rigs (in my sig) I was able to pull of some pretty high end OCs for air cooling, so I don't think the room or case temp is the issue here. I've tried reseating the heatsink 3 times after the initial install, twice with arctic silver and twice with the TX2 paste that came with the CCF. I noticed a drop of about 2-3c when using the arctic silver, so I'll stick with that. Everytime I've tried a different method for the paste, I've read some pretty interesting things about the direct core contact heatsinks and how to apply the paste. All of them result with the same temps, and after taking the sink off I notice that the paste is making good contact across all the heatpipes.

I suppose I could just be unlucky and got a very poor CPU, but I'd like to think that maybe I'm doing something wrong. I've tried turning up the CPU-NB multiplier (following dolk's amazing guide) and voltage, also I've tried the regular NB voltage too. I've tried lowering the HTT multi and raising it. I've tried the CPU multi method as well as FSB OCing. Also, just to get this out there, I ran memtest on my memory at its listed settings over night last night and it was flawless. So I think I can cross that off the list.

If anyone has any advice or suggestions, I'd very much like to hear it.

Thanks :)
 
Since the CPU has issues at stock speeds, if it was me that was having this problem,
I would probabily try a different PSU first, then maybe other components next.
 
MSI has a plethora of bioses out for that mobo, download and flash to one of the 25 and im sure you can find one that halfass works.

The only other thing that i see that could be causing you problems is the high voltage on your ram. High ram voltage causes heavy IMC stress so loosen the timings drop it to 1333 speed and drop the voltage to 1.65 and see if that helps your temps and your stability.
 
you said this

At stock settings I'm seeing about 35C idle and 50C load

Is it really at stock? whats your VCORE and northbridge and everything else at in the BIOS? can you give us a screenshot with CPU z along with core temp? I had temp issues with my BE until I went from 1600 to 1333 on the dram which allowed me to considerably lower all voltages.
 
After reading Weltall's thread, I set my memory to 1.7v and 1333 and currently running prime. I'll get to check after my 4 hour long Computer Concepts class.

I just learned how many kb are in a mb....sigh.
 
Cool stuff man. I just picked up that motherboard and Im learning lots about it.
A few things to start off:
Set regulart NB voltage to 1.3.
disable spread spectrum
make sure acc is on auto
disable CnQ
What is your NB speed? Start by getting the memory and CPUNB stable. Run prime for 5-7hours to verify.
For a 2600mhz NB on my 720 I needed 1.31 volts
If you cannot get stable at stock CPU core speeds with an OC'd CPUNB and memory at the 1333 settings, you can try setting PCIe voltage to 1.4 and the CPU PLL voltage to 2.6-2.7
Then slowy crank up your multi, bumping voltage as necessary. Run prime for 4-5hours before you make the next bump. It is a painstakingly long process and you just have to keep trying and trying different settings, even if you dont think they will change anything.
Not sure whats up with your heat issues. are you using the stock 90CFM fan on there?
It kept my 720 at about 55C running 4 cores at 1.5125 volts in 75F ambient temps, as a reference point for you.
 
I'd disable ACC for him since he has a 955BE. When I enable ACC on my mobo it doesn't let me boot up. I guess this is since the ACC is built onto the cpu with the phenom II's... don't know why it does what it does on the 720s though.
 
After switching my RAM to 1333mhz and 1.7v I ran prime 95 at 3.6ghz stable at 1.4v on the cpu, 1.25v on the NB-CPU and 1.2v on the NB for about 6 hours and maxed out at 54C. I was having trouble at 3.4ghz before at the same settings, so I'm thinking the memory is what was causing my stability issues. I'm going to drop the voltages a little bit more and see where I get with the heat. Then I'll mess around with RAM timings. Thanks for all the advice and info guys.

Also, I have spead spectrum, C1E support and CnQ all disabled already.
 
It's supposed to run 1.9v at 1600 and 1.5v for 1066. So at 1333 I just shot for 1.7 since it was in the middle. I know thats not very scientific, but so far it seems to be passing the blend tests.
 
At stock settings I'm seeing about 35C idle and 50C load, which appears kind of hot for the heatsink I have.
Definetly, thats really strange. I have a Xigmatek Dark Knight and Im usually getting ~28c idle, ~40c load.

I havent done much overclocking, I have the same motherboard you have. And from just some OC'ing I got it to 3.8GHz easily, didn't prime it for really long or anything but it seemed very stable through renders and games.
 
1.7V for RAM seems low to me. Though I'm on DDR2, I'm running 2.05V.

He's running DDR3 (1.5volt normal spec), that's why.

It's supposed to run 1.9v at 1600 and 1.5v for 1066. So at 1333 I just shot for 1.7 since it was in the middle. I know thats not very scientific, but so far it seems to be passing the blend tests.

I didn't notice the 1.9v in your original post. I'm not sure what these
were meant for, but I would go no higher than 1.65 volt with any CPU with
an IMC. I'm surprised you didn't fry your chip at 1.9v. :(

Edit

Good call to the poster who first pointed out the memory voltage issue.


1.65, btw, is +10% over 1.5 volts. Probabily should never increase any
voltage by more than 10% over the spec. While it's not guaranteed to
always be safe, designers usually try to build in a +/- 10% allowable range
on devices when they can.
 
I wasn't aware that 1.9v was too high, it's what the manufacturer has them rated at to run 7-7-7-20 2T at 1600 mhz. Right now I have it at 1.65v at 6-7-7-20 1T and it's priming just fine. Its' jumping between 50-55C loads at 3.6ghz (1.4v) depending on my room temp. I think I'll end up leaving it here for a while and just go for tighter RAM timings and be done with it. I"m not very comfortable going past 55C.
 
I wasn't aware that 1.9v was too high, it's what the manufacturer has them rated at to run 7-7-7-20 2T at 1600 mhz. Right now I have it at 1.65v at 6-7-7-20 1T and it's priming just fine. Its' jumping between 50-55C loads at 3.6ghz (1.4v) depending on my room temp. I think I'll end up leaving it here for a while and just go for tighter RAM timings and be done with it. I"m not very comfortable going past 55C.

I'm the same past 55C just kinda bugged me like i felt I was pushing the limits, backing off the memory speed let me lower all my voltages enough to drop a considerable amount in temps which let me overclock the CPU more which when benchmarking gives me better scores. The AMD's and 1600 I don't think really was meant to be without alot of voltage, If I had watercooling or something then I would probaby do it again at 1600.
 
Back