Home

Zalman VF1000 LED Graphics Card Cooler

Submitted by Mike Chin on Sun, 2007-11-04 23:51. Cooling
Viewing page 4 of 5 pages. Previous 1 2 3 4 5 Next

MEASURED TEST RESULTS

A. Stock Cooler

The first step in our testing was to see how the X1950XTX's stock cooler performed. It consists of a copper base connected to a group of fins by only a single heatpipe. The fan acts as a blower, pushing air over the fins out the exhaust grill. The unit is covered with a plastic shroud which prevents air from being expelled elsewhere.


The X1950XTX and stock cooler.

The stock cooler's underside.



ATI Tool's Fan Control settings.

According to ATI Tool, the fan ramps up in speed depending on the temperature reported by the thermal sensor. We were unsure what fan speed to use so we did a bit of informal initial testing. When left on default settings, the fan ran at 12% speed when idle and on load, the temperature stabilized in the low 80's with the fan alternating between 38% and 56% speed. 56% was very loud and at 38% the load temperature rose above 95°C, so to compromise we set it at 43% for testing. At this setting it wasn't loud enough for us to want to tear our hair out and the load temperature was more acceptable. We tried to set it at 40% but apparently the next step above 38% it would accept was 43%.

X1950XTX Stock Cooler Test Results
State
Fan Speed
GPU Temp
VGA Ambient
CPU Temp
AC Power
System Noise @ 1m
ATI Tool
43%
85°C
67°C
40°C
253W
32 dBA
Idle
12%
71°C
55°C
25°C
130W
22 dBA

At 43% the noise signature was smooth but overall was quite loud. Anything higher sounded like a small engine or vacuum. At 13% the fan emitted a slight but noticeable hum. At this level the test system registered at 22 dBA which was rather low, but the type of noise was unpleasant. The stock cooler cooled the card admirably, though at a noise level few could tolerate. Let's see what the VF1000 can do.

B. Zalman VF1000 LED

Zalman VF1000 LED Test Results
State
Fan Speed
GPU Temp
VGA Ambient
CPU Temp
AC Power
System Noise @ 1m
ATI Tool
12V
73°C
54°C
42°C
248W
30 dBA
ATI Tool
9V
76°C
56°C
43°C
249W
28 dBA
ATI Tool
7V
78°C
57°C
43°C
248W
24 dBA
ATI Tool
5V
79°C
58°C
43°C
248W
21 dBA
Idle
5V
46°C
39°C
26°C
124W
21 dBA
Note: the stock memory heatsink was used

@12V: The noise was very high pitched and whiney. Cooling performance was incredible with both the GPU and surrounding environment a good dozen degrees lower than the stock cooler.

@9V: The pitch of the overall sound dropped somewhat but high turbulence kept the noise level intolerable. Performance dropped off by a few degrees.

@7V: The fan exhibited some unpleasant undertones but it was hard to detect as turbulence was dominant. At 2°C higher than 9V, the GPU was still fairly cool.

@5V: At 1m, the fan was basically indistinguishable from the rest of the noise emanating from the system. Performance was still very good at 79°C — only 1°C higher than at 7V. In addition it was much quieter than the stock heatsink yet 6°C cooler.

Overall, these are very impressive results with only 6°C separating the 5V temperatures from the 12V temperatures. At 5V it's clear the VF1000 cooled the X1950XTX better than the stock cooler with only a small amount of airflow without adding any noise to our already quiet test system. System power consumption also dropped a few watts, possibly because the video card fan was powered externally. The older VF900-CU has a tough act to follow.



Previous 1 2 3 4 5 Next

Cooling - Article Index
Help support this site, buy from one of our affiliate retailers!
Search:         
» 315571 reads