FSB Wall Behavior???

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Using the hardware in my signature I have run into a bit of a snag with my OC. I know I'm still using the stock cooler, but I have removed the stock paste and applied some of my own (not AS5, but still a step up) and re-seated the cooler to make sure it is making proper contact. What happens is that even though I'm good for 10 hours of Small FFT's at the clock/voltage in my sig, 3.3GHz /w a small bump in CPU voltage results in the system basically just shutting down after a few minutes. The shutdown temperature in the BIOS is disabled and I'm only at about 65C load at the time of shutdown.

I haven't checked my NB temps but at this point I figure it's either NB temps, NB/SB voltages, or an FSB wall.

I'm not sure if a "random" shutdown like that would be characteristic of an FSB wall though, and I'm more or less looking for a bit of guidance in figuring out what's causing the shutdowns. I've tried more Vcore...but it actually makes is shutdown sooner, which kinda makes me think that even though I'm only at 65C load, that this particular CPU might need to be cooler then that at 3.3GHz.

What do you guys think?
 

luihed

Limp Gawd
Joined
Feb 15, 2007
Messages
510
A fsb wall means your cpu will not go over a certain fsb regardless of multiplier used..... My cpu has a fsb wall at around 555fsb, at first I thought it was my motherboards limit till I put my friends E6300 and his cpu reach 566fsb ..... Doesnt matter if I use 8x, 7x or 6x , I cant get over that fsb wall ( i did reach 556fsb once:) .....
 

Xilikon

[H]ard|DCer of the Year 2008
Joined
Oct 12, 2004
Messages
14,845
My own fsb wall experience was with a recent attempt with a E6420. No matter what, it is perfectly stable with orthos @ 376 MHz but at 377 MHz, the computer won't even post at all. I know it's not the motherboard because I tried to run my E4300 and it ran fine @ 400 MHz. Turn out the CPU is defective since with this wall, it's just 3 GHz, no a mhz extra with the 2 core temps being 3-6C apart :eek:

Yours isn't really a fsb wall but more like stability issue. Try to feed more voltage to the cpu and see if you can push it further. With a wall, voltage doesn't matter at all.

 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
My own fsb wall experience was with a recent attempt with a E6420. No matter what, it is perfectly stable with orthos @ 376 MHz but at 377 MHz, the computer won't even post at all. I know it's not the motherboard because I tried to run my E4300 and it ran fine @ 400 MHz. Turn out the CPU is defective since with this wall, it's just 3 GHz, no a mhz extra with the 2 core temps being 3-6C apart :eek:

Yours isn't really a fsb wall but more like stability issue. Try to feed more voltage to the cpu and see if you can push it further. With a wall, voltage doesn't matter at all.


That's what I was thinking, because it posts and sits in windows as long as I'm not stress testing.

More vcore makes it fail sooner, so it might meen that it is a temp issue after all, I guess it's possible that this chips needs to be cooler then 65C to do 3.3GHz.

I'll keep looking into it, but I guess if it comes down to needing more then stock cooling I'll HAVE to get something else. :p

A fsb wall means your cpu will not go over a certain fsb regardless of multiplier used

I know what an FSB wall is...but I wasn't sure if some CPU's exhibited different behavior, as I had never worked through a wall myself. Thanks, but i should have made that clearer.
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
haha...I think I might know what it is...while dual folding my chipset is at 70C...and Small FFTs will push it even farther, aside from the added stress of running at the higher bus speed.

Any thoughts as to whether or not a chipset temp of 70C-75C could be causing this board to "buckle" and not handle the higher OC as well?
 

silent-circuit

[H]F Junkie
Joined
Sep 18, 2005
Messages
16,133
Yes, I'm fairly sure that's too high by at least 10C to ensure normal, consistent operation. You need to increase cooling on it somehow, maybe try backing off voltage to it if possible (possible on a 680i, not sure about others)
 

Xilikon

[H]ard|DCer of the Year 2008
Joined
Oct 12, 2004
Messages
14,845
70C is already pretty high so it's normal it will get unstable if you push a bit. You may need to get a better cooler to reduce the temps. the stock Intel cooler is able to sustain up to 3.2 GHz. Anything over this and it will be unable to dissipate heat fast enough to keep the temps in check.

You will be better using a aftermarket coolet ;)

 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
I need both the HR-05 and the Ultra-120 at this point..hopefully it's not long before they come my way.
 

Slammin

Limp Gawd
Joined
Jan 30, 2005
Messages
171
Crazy thread.

60C is not too hot for a 680i mobo chipset though some don't like them that high. My bios default for shutdown due to PWN temp is 125C and my PWM sets at around 45-48C depending on ambient, with my OC. At load, my PWM reaches 80C.

I think some in this thread are confusing which temp you are having the problem with, but if it is chipset, you should check to see what is normal for your mobo, but I would say that is way too high considering you are only running a dual core cpu, so chances are it is spiking much higher than 70C but you can't see that since it shuts down. Bad chipset cooling sounds like the culprit to me.
 
Top