Issue with an Opteron 165

[mgpaulus]

I picked up a Socket 939 Opteron 165 off ebay, and I seem to be having an issue with it, in that I can't seem to get it stable like I would expect. I have the chip in a Gigabyte GA-K8NS-939 board. It came with some stock cooler, but I would seem to get thermal runaway and shutdown with that cooler within 5 minutes. So, then I ordered one of the 4 heatpipe Opteron stock coolers, and that allows the CPU to stabilize around 50C, according to the BIOS monitoring. However, as soon as I kick in the Mersienne Prime testing from UBCD, I again achieve thermal runaway and shutdown.

I was hoping to replace an Opteron 148 with this and up my crunching power a bit, but it doesn't seem to be wanting to get stable.

Do I have a bad/hot chip, or am I missing something? I have 3 of these coolers on some 3800+ X2 chips, and they seem to work fine.

 

[VirtualLarry]

The 4-heatpipe stock cooler should be enough for that chip. I had one that I overclocked to 2.5Ghz stable, and I had an X2 3800+ that I overclocked to 2.6 or slightly higher stable on that cooler as well.

What type of thermal grease are you using?

Edit: You say it hits 50C, before PRime95 testing, does that mean at idle? Something is very wrong if you are at 50C at idle. You should be around 36C or so.

 

[NXIL]

I have an Opteron 170 with that stock heatsink: currently, 81F, 27C.

Either: some mechanical problem, i.e. heatsink not making good contact, thermal interface material is too thick or too thin (? too thick is more likely from what I have seen), chip itself bent or not flat, some problem with heatsink attachment points.

Next: fins on cooler dirty? Fan working?

Cool and Quiet set incorrectly?

Mobo having issues running the fan on the heatsink?

You got a new cooler, so fan itself should be OK.....

Oh: voltage set correctly in BIOS?

Can't think of anything else, and, can't think of a reason why the chip itself would run hot.

GL

NX

 

[cubeless]

i wonder if some of these chips are starting to leak current after years of being oc'd and ov'd... boxes that i built with x2 3800 that have been @ 1.5ish v and 2.5ghz for years now are running hotter than they used to... new as5 doesn't help...

 

[AznAnarchy99]

My old 3800x2 was stable at 2.6 for years until recently it would lock up on games when lots of shit was going on. Dropped it down to 2.5 and it was all good after that..

 

[Smoove910]

<---- wonders if it needs de-lidded...

I've heard/read stories from back in the day where some of the TIM was not spread well enough between the actual die and the bottom of the heat spreader on the processor itself.

I delidded my 148, but was only warranted with a couple degree drop in temps. However, if the interface material between the heat spreader and the actual core is shitty, there's a good chance that it could be to the point where it always runs run (not making actual contact).

If you decide to de-lid, be extremely careful since there's a good chance of shaving off a couple resistors and/or cracking the die.

Good luck either way!

 

[robertk2012]

You should have bought my chip.

Make sure your not using to much AS5 or whatever your using. The AMD Heatpipe cooler should be more than enough for that chip. How is the airflow in the case?

 

[mgpaulus]

Thanks for all the input guys. I'm going to look closely at it. I'm wondering if it was delidded before maybe and the lid got somewhat whanked. I put on a thin smear of AS5 all over, and when I pulled the HSF off, all I see is a ring around the HSF, where it looks like the edge of the CPU made contact only, and not the inner part.

Right now the whole rig is on a bench in the basement, so there are no airflow restrictions.

 

[robertk2012]

Dont smear it on. Put a small amount in the center.

 

[robertk2012]

Any luck?

 

[bodoblues]

I had almost the same problem with an 165 on an a8n sli se, although same processor worked fine on a8n sli deluxe. I solved the problem by decreasing the vcore to 1.1v and it become stable even overclocked nearly to 300 fsb. I guess it depends on the stepping on some boards. now I have another 165 on the a8nsli se and it's running like it should vithout decreasing the vcore. good luck with yours!

 

[theevilsharpie]

Dont smear it on. Put a small amount in the center.

This.

 

[zmatt]

Use TX-2 not AS5. Make sure you have good airflow. Check your BIOS settings (vcore etc). Is the fan spinning? Your temp diode may be bad also. Check the temp from other things, video cards, mobo, hard drives. Are they within normal temps or are they overheating too?

back in the day I had an Opty 165. It overheated at first, but that was because i couldn't afford to get a real case. I had to run it with the side panel off all the time. When I got a better case with front and rear fans temps dropped and i could get it to 2.4ghz on air very easily.

 

[PiTTeRaTTeR]

i have a oppy 165. had it on a abit atx-32x. it ran stable with stock cooler at 2.8 with very slight volt increase. temps were nver over 50.