Glaciator C/W results

[Len]

For anyone interested here are some C/W values I arrived at for my Glaciator cooling a Tbird 1200 AVIA running at either 1.85v or 1.95v. I used the 'CPU Stability Test 6.0 (CPU warming only)' program for 10+ mins to make the CPU reach max temp. Temp reading are from MBM5 using the standard built-in mb temp sensors.

There is a possible trend of the Glaciator getting less efficient as the cooling work it has to do increases. But overall it seems to provide excellent cooling.

1000 Mhz @ 1.85v
C/W = (43C load - 33C case) / 60 watts
= 10C / 60 watts
= .17

1333 Mhz (133 X 10) @ 1.95v
C/W = (53C load - 35C case) / 85 watts
= 18C / 85 watts
= .21

1380 Mhz (138 X 10) @ 1.95v (Actual test)
C/W = (54C load - 36C case) / 88 watts
= 18C / 88 watts
= .20

1400 Mhz (140 X 10) @ 1.95v (Actual test)
C/W = (55C load - 36C case) / 89 watts
= 19C / 89 watts
= .21

 

[Phil]

So you were using an in socket thermister?
If so then those are probabally better results than they seem.

 

[Len]

Whatever is on my Epox 8K7A motherboard. The results are good but my 1200 AVIA Tbird seems to still top out around 1430 or so at 1.95v. I backed it off a bit for a little stability tolerance. My next step is to drop from 10.0x to 9.5x or 9.0x and push the FSB to 150+.

 

[Hoot]

Get those temps down in the 30-40C range and it'll go faster. See my sig. That's not as fast as my Tbird will go, but a good compromise against system strain.

Hoot