AMd 4800 X2 temp 95c + under load "ADO4800DDBOX"

Greetings i have an interesting problem with my new AMD 4800 x2 cpu.
when using the stock cooler the cpu temp raises to 95c or more, the system shutsdown, this could happen during a normal windows install.
Aha! i thaught i have seen this before its the termal paste that is not good enough so i got some arctic silver 5, same thing happens, computer shuts down under windows install. Ok i think i get a big cooler for it so i get a zerotherm 95, and install some more chasis fans. Now i can install windows, but if i run cpuBurn, the temp quickly reaches 85 C.
I used what would amount to around a matchheads worth of cooling paste, and the surfaces of both cooler and cpu had been cleaned.

Idle temps on the cpu are 36C which is below what the MB is at which is 37C.

Can anyone tell me what i do wrong, in my lille world the stock cooler should be able to handle the cpu that it is delivered with?

the motherboard is a asus M2N32-sli deluxe wireless edition.

what i do see is that CPUZ lists the cpu 2.5 Ghz which i feel is right but the motherboard tells me that the cpu is 2.4 GHz, no where in the bios have i found the 12.5 multiplier.

the vcore is 1.36 v could this have something to do with it?

br
Lasse Knudsen

Weird, I have the asus crosshair (slightly more expensive gaming version of the m2n32 sli) and that hasn't given me any temp problems like that. Try leaving the mult at auto and see if that will set it to 12.5

Also, try reseating the hsf and check if it's not perfectly balanced on there, your hsf may need lapping (btw, is it the heatpipe or regular one?)

It seems unlikely that you have an unsually hot cpu, but it's possible, something you might want to try is setting the mult at 10 and manually reaching 2.5ghz (or higher if you plan to oc) and see if the 12.5mult is retarded in some weird way

But if you do plan to buy a new hsf, msn me and I'll help you there in customizing your cooling

If your CPU ever did 95c it wouldnt work no way. Your mobo must be readying the CPU wrong. Try updating your bios. Many mobo's read the brisbane wrong. My mobo with the latest bios update locks the temp at 40c. Using speedfan to check my temps say -4c which isnt correct.

Many ASUS mobo users say their idle temps are 60c with brisbane. My mobo has a shutdown temp setting which your mobo may have. 1.36v is to high as mine OC'ed to 2.51GHz only needs 1.28v. Sounds to me like your mobo thinks your brisbane is a 90nm. Your x2 4800 should work below 1.3v so try this voltage and keep the voltage low as possible while stable. Until ASUS gets off their behind and updates your bios with the last update over 4 months ago your kind of stuck.

The only two things I can recommend is re-seating the HSF again with a very thin layer of AS5 and equipping the CPU heatsink with a fan of it's own (blowing in the direction of your exhaust fans); you can put a 120mm fan on it with screws, because your current configuration as a passive cooler is good only when you have a low power CPU but for overclocking and overcoming issues like yours, these kind of coolers are always equipped with fan.
This last move will bring your temps down by another 5-15°C, enough to install windows properly and then check the Vcore with CPUZ, because most probably, the real problem is with your motherboard, which might me overvolting the CPU and causing it to heat up that much, because the idle temps (when C'n'Q sends the CPU on 1.0GHz/1.0V) are OK and do not suggest a particularly hot chip.

Greetings and thanks for all the replies

the current situation is that i have reversed the chasis fans, and i have reduced the vcore to 1.23v but im still at 72 degrees when i load the system with 2 intstances of cpu burn.

to my best knowledge all the heatsinks i have tried have been seated tightly, i have not been able to move them, infact even after i remove the clamps, it takes some force to remove the heatsink.

I have no idea how low i can put the vcore without having any trouble.

and the only bios i can find on the asus webpage are beta ones, and i cant see that they do something about my cpu

br
lasse knudsen

As mentioned, download the latest BIOS - try the BETA and see if it helps

Also, have you touched the HS while the CPU is under load? If it's not getting warm, then it's either a BIOS issue or bad contact (I highly doubt the bad contact between the IHS and HS as your extensive troubleshooting would've solved that issue).

However, in an unlikely case, the IHS may not be properly seated, causing poor thermal dissipation.

The heatsink,... at the end, it's the only thing that makes sense to explain such problem, however, never had a problem like this; today just upgraded my girlfriend's sempron with a 65W (90nm) X2 3800+ and, at least temperature-wise it was the end of the world; 30°C idle/41°C burn in a 21°C room.... It seems to me so outlandish that a X2 can go 95°C with the HSF mounted and working properly,... a defectively placed heatspreader can be the only answer. If he is careful enough, yes, he can remove it and place the heatsink directly on the core like we used to do on most PIIIs and AthlonXPs but today's heatsinks are clamped harder than the previous and the risk of damaging the core is much higher; in this case I'd take kind of a Zalman 7000AlCu or any other cooler that has it's own mounting plate, so I could control the overall pressure of the heatsink on the core.

I suppose that's logical, but amd started soldering the ihs on all 65nm amd cpus, so it would be very difficult to take it off without breaking the cpu. why doesn't he just send it to amd for an exchange if the heat spreader's defective, I doubt they would refuse an rma for that reason

asus q-fan (or ai quiet) in the bios - disable it, and make sure your hsf is seated properly (including nothing around the cpu area that could lift the hsf off the cpu)

Thats just par for the course.

I prefer being a normal OCer. I'm still trying to figure out a way to safely do that mod. If I can't, f*ck it. Otherwise I'll go bother Freecableguy on how he did it.

It turned out that it killed the chip. No focking go. Thank God for spare Celerons.

I'll send it now.