One core stays about 8-10c hotter than the other...

ny_driver

Registered

​

Joined

Dec 15, 2009

Location

central new york

• Dec 26, 2009

• #1

On my other Opteron 170 the temps were pretty even, but on this one I get one core way hotter. Could it be a faulty or misplaced sensor? The lower of the 2 cores is at a very acceptable temp range idle and load @ 50% OC, but the other core is in the low 60's @ load. I ordered a sunbeam core contact cooler to replace my AC Freezer 64 pro which should help all around, but the 10c difference in core temps still troubles me.

 

]-[itman

​

Joined

Sep 24, 2001

Location

Arizona

• Dec 26, 2009

• #2

Leakage problem on one core? Do they run very different temps at stock load?

 

Q

QuietIce

Disabled

Joined

May 7, 2006

Location

Anywhere but there

• Dec 27, 2009

• #3

The temp difference you see, although near the acceptable limit, is normal. The K8 internal sensors are not tested at the factory so there's no offset burnt into them. In fact there is no supported internal sensor at all (even though it's there and we use them all the time). I've got five running s939 systems and they vary from a 1-2°C difference to a 5-7°C difference and I've read about them being as high as yours - but never over 10°C. I wouldn't worry about it ...

 

OP

ny_driver

Registered

​

Joined

Dec 15, 2009

Location

central new york

• Dec 27, 2009

• Thread Starter
• #4

I killed it popping the IHS off to put new TIM inside. Now I need to find a good deal on another comparable chip.

 

Bluefalcon13

Member

​

Joined

Jan 1, 2008

Location

Colorado Springs, CO

• Dec 27, 2009

• #5

how did you kill it? I've de-lidded 3 AM2 (one windsor core, and two barton cores) without any issues

 

OP

ny_driver

Registered

​

Joined

Dec 15, 2009

Location

central new york

• Dec 27, 2009

• Thread Starter
• #6

I tried prying before I had proper instruction and dug a scratch in the the little black rectangle along the edge under the adhesive. Now all I get is beeps. Very stupid of me.

 

OP

ny_driver

Registered

​

Joined

Dec 15, 2009

Location

central new york

• Dec 29, 2009

• Thread Starter
• #7

The Opty lives. I took out one stick of RAM and voila! But now I can't get it to POST with 2 sticks of RAM, even tried a different pair. All the RAM is memtested, and works fine. Any ideas why it is acting this way?

 

Q

QuietIce

Disabled

Joined

May 7, 2006

Location

Anywhere but there

• Dec 29, 2009

• #8

Has a known good stick been Memtested in each slot by itself? How did that turn out ...?

 

OP

ny_driver

Registered

​

Joined

Dec 15, 2009

Location

central new york

• Dec 30, 2009

• Thread Starter
• #9

QuietIce said:

Has a known good stick been Memtested in each slot by itself? How did that turn out ...?

Click to expand...

you can't test slot 2 or 4 by itself, however the RAM has been tested several times recently, 2 sticks together...both pairs(2x1024 OCZ platinum, 2x512 Crucial Ballistix). Anyways...I tried the cpu in a different mobo and I get the same results(can't run in dual channel)

 

Q

QuietIce

Disabled

Joined

May 7, 2006

Location

Anywhere but there

• Dec 30, 2009

• #10

Doesn't sound good ...

 

OP

ny_driver

Registered

​

Joined

Dec 15, 2009

Location

central new york

• Dec 30, 2009

• Thread Starter
• #11

yeah it sure doesn't. lesson learned....(the hard way as usual) I bought a 4400+ on ebay for $75+$6 shipping. the ad said $100 or best offer...I think I got a great deal considering what they're going for these days. They have at least 20 more of them for sale.
I only paid $80 for the Opteron...I've got to test it and see what it'll do in single channel mode. It did 3000 MHz no problem in dual channel with tight memory timings 3-3-3-6-1T.
http://products.amd.com/en-us/Deskt...SOI&f8=&f9=2000&f10=False&f11=False&f12=False