Specification

The characteristics of the new graphics card cooler from Zalman are listed in the following table in comparison with the last-generation Zalman VF700-Cu model:

You can notice that the maximum fan speed of the new cooler is considerably lower than that of its predecessor. It means the level of noise is lower, too. Besides the heatsink design, the Zalman VF900-Cu LED has the following advantages over the VF700-Cu model: a Fan Mate 2 speed controller included into the package, fan highlighting, and thickness (15mm against the VF700’s 10mm). Alas, these advantages come at a price, which is almost two times the price of the Zalman VF700-Cu.

Testbed and Methods

We tested the Zalman VF900-Cu LED and its today’s opponents on the following testbed:

• ABIT AN8 SLI mainboard (nForce 4 SLI chipset, Socket 939, BIOS v.2.0)
• AMD Athlon 64 3200+ CPU overclocked to 2800MHz (2000MHz default clock rate, 1.625V, 512KB L2 cache, Cool&Quiet Disabled, Venice core, E6 stepping)
• 2 x 512MB Corsair TWINXP1024-3200C2 DDR SDRAM
• Graphics cards:
• Sapphire Radeon X800 GTO2 (256MB GDDR3, 540/1240MHz, 16 pixel processors)
• Radeon X1900 XTX 512MB (650/1550MHz)
• Maxtor DiamondMax 10 hard disk drive (6V250F0, SATA-II, 250GB storage capacity, 7200rpm, 16MB cache)
• System case: ASUS ASCOT 6AR2-B Black&Silver + an intake 120mm system fan Coolink SWiF (~1200rpm, ~24dBA) + an exhaust 120mm system fan Sharkoon Luminous Blue LED (~1000rpm, ~21dBA)
• Power supply: MGE Magnum 500 (500W) + 80mm GlacialTech SilentBlade fan (~1700 RPM, ~19dBA)

I tested the cooler on two graphics cards: 1) on a Radeon X800 GTO2 which has a moderate appetite by today’s standards, but I enabled 16 pixel pipelines in it and overclocked it a little above the default frequencies of the Radeon X850 XT PE, and 2) on a Radeon X1900 XTX which is unrivaled in terms of heat dissipation among today’s graphics cards. So, you will see if it makes sense to install the Zalman VF900-Cu LED on modern and high-power devices as well as on graphics cards with relatively low power consumption.

The tests were performed in Windows XP Professional Edition Service Pack 2. The ambient temperature remained at 25-25°C during the tests.

The unblocked and overclocked Radeon X800 GTO2 had to run the Firefly Forest test from 3DMark2006 for ten times without full-screen antialiasing but with 16x anisotropic filtering. The Radeon X1900 XTX was heated with the Deep Freeze test with the same settings. The temperatures were monitored through RivaTuner v2.0 RC16. Each cooler was tested at least two times, and the temperature stabilization time between the test cycles was 30-40 minutes.

I took the following coolers as opponents to the Zalman VF900-Cu LED in this test session:

• The stock coolers of Radeon X800 GTO2 and Radeon X1900 XTX cards
• Zalman VF700-Cu, one of the best graphics card coolers today
• CoolerMaster’s Aquagate liquid cooling system with the Va-Aqua water-block (RL-VNC-CNU1, for the X1900 XTX only)

The stock coolers were tested in two modes: when the fan speed was controlled automatically and when it was manually set at the maximum. The Zalman VF700-Cu was tested in the quiet mode (~1980rpm) and at maximum fan speed (~2820rpm). The Zalman VF900-Cu LED was tested in three modes: 1) silent or at min fan speed (~1320rpm), 2) quiet (~1980rpm) and 3) max fan speed (~2400).