Add to Reddit
Heatsink Compatibility
Due to the rather unique design, Zalman cannot guarantee that it will fit all graphics cards. But luckily, all the popular cards are supported, including all ATI and NVIDIA GeForce graphics cards, as long as there are two mounting holes on the card.
Matrox and NVIDIA's PCX cards are not supported because of the odd angles of the mounting holes and the location of the GPU.
){kind=logo}
Logo
According to Zalman, the ZM80D-HP (old heatpipe cooler) will not fit on the NVIDIA GeForce 6800 series because the mounting holes are only 2mm wide, while the Zalman screws are 3mm wide. Fortunately, VF700 series will fit a GeForce 6800 series cards just fine. This is a big plus point over the ZM80D-HP.
)
Mounted
I won't be covering heatsink installation since it's 100% identical to VF700-AlCu version. If you want to read about its installation, please click here.
Test Bed
| Test Bed |
|
| Processors |
AMD Sempron 3100+ Socket 754 |
| Motherboards |
ASUS K8V Deluxe |
| Memory Modules |
1GB Corsair PC3200XL TwinX |
| Graphics
Card |
Sapphire Radeon 9800 Pro 256MB (modded to XT, running at 380/340) |
| Sound Card |
EMU 0404 |
| Power Supply
Unit |
Zalman ZM400B-APS |
Testing Methodology
In this review, I will be concentrating on getting consistent results, which is rather difficult on a graphics card. I improved the testing methodology to reduce inconsistency - instead of measuring room temperature, I placed the thermal probe near the fan intake area.
The thermal probe measuring the GPU was placed at the same spot as previous reviews. Because of this, the results here are not comparable with any temperature readings from past reviews. Arctic Silver Lumiere, a special testing compound was used and reapplied before every installation.
To check for consistency, I recorded the temperature 3 times per heatsink, and tested both 5V and 12V settings each time. The stock heatsink will be included in this roundup, but only in the 12V mode.
We used the Omega HH506R thermometer, which supports two thermocouples and RS-232 logging functions. We used two Type-K thermocouples for these tests.
I switched to Rthdribl v1.2 (Real-Time High-Dynamic Range Image-Based Lighting) from Aquamark 3 for better consistency. Testing time remained the same, at 10 minutes per test. Case and GPU temperatures were recorded simultaneously using a secondary PC via RS-232.
RAM heatsinks were not installed during testing because the stock cooler wouldn't fit if the RAM heatsinks were installed.
