Hey,

I used to have an AMD X2 4200 processor and it was running fairly hot - around the 50's (°C ) idle - I think this may have been causing a random rebooting problem I was having. Anyway, I just upgraded to a 5600 and it's running even hotter. It idles at around 60°C and breaks the 70°C mark when under load. My random rebooting has also increased in frequency from once every 1-2 days to 3 times today! I thought my cooling was okay - I have an Antec gaming case with 4 fans (2x120mm in, 1x120mm and 1x240mm out) and I don't have overheating problems with other parts like the graphics card.

The stock cooler goes at 3000rpm. Is this the highest it will go? I'm sure I've read that it can go higher, but I can't seem to adjust it - I've tried installing SpeedFan, but it will only let me adjust it down (and that doesn't seem to have any effect). My motherboard is an Asus M2N4-SLI. I've also tried Motherboard Monitor, but it doesn't even recognise my sensors. Cool & Quiet is on, but it's not helping at all. Any ideas?

Thanks,

-Joe

What kind of thermal paste/adhesive are you using between CPU and heatsink?

The stock AMD heat sink came with thermal paste pre-applied, just like the last one. It was just a matter of plugging in the processor and clicking the heat sink in on top of it.

Dannnnnnggggg

X2 4200+ @ 50C and X2 5600+ @ 60C idle..... WOW..... Makes comparison with a early gen P4 PRESCOTT nothing when looking at thermal issues.

Well... I got some suggestions.....

1. reset CMOS - get everything in bios back to default. there could possibility that Vcore could be set incorrectly as in higher voltage.. by reseting CMOS, this should set all settings back to there default.

2. update bios - by updating to the latest bios, there could be a fix for temp readings to an earlier bios where there could be an incorrect reading of the sensor and a bios update could fix that.

3. Get some more thermal paste and reapply after cleaning the old stuff off.... with this, this can give you an opportunity to make sure the sink is seating properly on the cpu.

I already tried resetting the CMOS and updating the BIOS to the latest version.. no joy.

It seems I've come to the same conclusion about it being the contact between the CPU and the heat sink - with it running at 60-70 degrees, the metal part of the heat sink under the fan should be hot to the touch, right? Mine is cool to the touch. Looks like I'm going to have to try a different thermal paste - thanks for the suggestion. Though it seems weird that it's not making proper contact when both the CPU and heat sink are brand new (arrived today) and I installed it as per the seemingly straightforward instructions.

In the mean time, I'd just like to use my PC without it rebooting all the time. What would be the best way to temporarily cool down the CPU, e.g. by reducing the voltage or underclocking it? Asus doesn't seem to be much for overclocking software, so I'm not sure what to use..

if the heatsink is not getting hot thermal paste aint going to make much difference.

is the heat sink contacting the heatspreader of the cpu

now this is just a hunch but maybe the bracket is defective on your mobo making poor contct with the cpu

also set the vcore fsb and multi manually incase its not right

I now know it's definitely contacting properly - when I opened it up to remove the heat sink, it was stuck onto the CPU quite well. I had to heat it up and twist it gradually to get it to come off. I've replaced the old paste with some Akasa pro-grade 460 thermal compound, which seems to have cooled it down somewhat - though it still peaks at an uncomfortable 65°C, that's below the specified maximum of 70°C, so I guess it's okay for now. I'm not sure CPU heat is actually causing the crashes now, but I may consider buying a decent CPU heatsink/fan anyway.

I tested it by running Half-Life 2 in windowed mode and looking at the temperature - it was at 65 in the main menu (which has that in-engine 3D backdrop), at which point my PC rebooted. I've also had a few blue screens - before when I opened up the PC to change the fan speeds it started to refuse to start up - I got some assorted blue screen messages on the first XP loading screen when trying to do so. From a quick search, they seem to point to either a CPU problem or defective memory. Strangely, it also froze up with I opened the BIOS menu - the clock still updated, but there was no keyboard feedback. In the end I had to reset the CMOS by taking out the battery.

Does this sound like it might actually be my RAM? Surely if the CPU is operating below specified limits then it really shouldn't be causing reboots, even if the temperature is less than ideal. Would memtest86 find any problems with RAM, or am I going to have to start swapping out sticks?

Hey buddy,

If you have a couple of sticks of ram, try one at a time. Chances are it is just one of the sticks, highly unlikely it is both.

Memtest would indeed show up any ram instabilities.

My favoured way of using memtest is one stick at a time,run the full test, then put in the other stick and run it again. It may take a while, but at least you'll know.

Either way, your temps are WAY too high for stock speeds.

If it's not contacting properly it could be badly concave/convex. Have you considered lapping it?

What do you mean by lapping? I don't see any major concavity or convexity in either surface..

Anyway, I've been looking at other monitoring programs and now I'm slightly confused. Nvidia Monitor and ASUS PC Probe (which came with my motherboard) both agree on the aforementioned CPU temperatures - but SpeedFan is ambiguous. Temp1, which seems to be the same value as the ones in Monitor and PC Probe, doesn't indicate what exactly is being measured - all it says is that the chip is an IT8712F-J. A different value labeled more specifically as "AMD K8 Core" seems to think my temperature is about 21C idle. Yet another program, Core Temp, gives the temperatures of both cores - 31C and 21C idle respectively. Which sensors should I actually be looking at?

Use the BIOS. If the BIOS says it idles at 40C and apps are saying 60C, the app is off by 20C. I have yet to encounter BIOS that reported the temperature incorrectly.

wont help much if you're testing load temps.

1 value (the lower one) is most likely the value between the cores, under the heatspreader. The other, slightly higher values are most likely the values in the cebtre of each core.

The 'chip' you refer to is the thermal probe inside the CPU.

I would imagine CoreTemps values are correct for each core, its the most reliable temp monitoring program in my experience.

my amd 5600+ is running at 30c overclocked to 3.0Ghz and the highest ive ever seen it get to is 45c with stock cooling

Same here. Joe, I can bet you don't have proper cooling, or have poor air ventilation in the case.

Look at my processor, it is overclocked to 3.20GHz WITH 1.38V and it stays as cool as 30Deg C idle and maximum 50-55Dec C at full load, and that too WITH STOCK HSF. Consider getting more fans for your case.

I just took a screenshot for you, click on the following link to see yourself:
http://img219.imageshack.us/img219/1395/18779573gn1.jpg

Since we're resurrecting this thread, I just want to confirm that I came to the conclusion that it was reading the temperature incorrectly. I still haven't determined the cause of the rebooting, but I think that it could be a software issue as I occasionally get blue screens - as I've checked or replaced everything from RAM to graphics card since I started getting the problem, these reboots could be something completely new...

I've checked Core Temp again, and it's still giving me more reasonable temperatures: ~38 and ~27 Celsius. I think I'll just go ahead and trust it on that.

Oh, and by the way..

Supreet Virdi said:

This is my case: http://www.diycomputers.com.au/images/upfile/2007291647024143.jpg

It has one 200mm fan at the top, one 120mm at the back and two 120mm at the front. I think there may be space for one or two more fans on the side, but I'm not sure that'd help any more..

hey i have had similar problems before in the past with random rebooting on many different pcs and ive had a few ! the problem i found was corrupt ram sticks, which confused me even more when i bought new ram one day and i had the same problem only to have confirmed by the supplier that the new stick was faulty if you have more than 1 stick i suggest you try running for a few days without one of them and see how you get on if it still fails swap sticks until you find the bad one short of that replace them all.