nVidia GeForce4 MX 440 in the test: Inno3D, Gainward, Leadtek and Prolink in comparison

nVidia GeForce4 MX 440 in the test: Inno3D, Gainward, Leadtek and Prolink in comparison

Test system

This time we tested a system based on a combination of SiS745 chipset and AthlonXP 1700+ with 256MB DDR-RAM is used. We tested under Windows2000, because it represents the basis for the home OS of the future in Microsoft's portfolio and therefore it is easy to estimate how reliable and high-performance the respective drivers will be under this operating system.

That Service Pack 2 and the current application updatefrom Microsoft were installed as well as chipset drivers of version 1.09 from SiS. The latest DirectX 8.1 goes without saying. The GeForce cards were tested with the recommended driver v27.30, for the KyroII, which is included for the sake of completeness, version 14.0028 of the official reference driver was used. For background noise, the subjectively satisfactory onboard sound of the SiS chipset with the 7012.1.03c drivers was used for a change.

  • CPU:
    • AMD AthlonXP 1700+
  • Motherboard:
    • Elitegroup K7S6A with SiS745 chipset
  • RAM:
    • 1 * 256MB Nanya/Apacer PC2100 DDR-RAM CL2.5
  • Graphics card:
    • Inno3D Tornado GeForce4 MX440
    • Gainward GeForce4 PowerPack pro600 TV Golden Sample
    • Leadtek WinFast A170DDR T
    • Prolink Pixelview GeForce4 MX440
    • Inno3D Tornado GeForce2Ti
    • Inno3D KyroII 4500 32MB
    • Asus v8200pure clocked at GeForce3 Ti200 level
  • Peripherals
    • IBM DTLA 307015
    • Pioneer DVD A03
    • 3Com 3C905-TX 10/100MBit network card
  • Other
    • A4 Tech Double Wheel Mouse
  • Software
    • Windows2000 Professional

So that the graphics are not too confusing, w earth the four GeForce4 MX440 cards shown together. In the test, there was no difference in performance beyond the usual measurement inaccuracies. The exception is the Gainward card, which the manufacturer delivers with a switchable overclocking guaranteed ex works.

The KyroII is mainly used to simulate older cards, many of which have to make do with 32MB. In no case should she serve as a sacrificial lamb here, as in the majority of the benchmarks she falls behind more or less significantly due to her age. To put the 32MB limitation into perspective,some settings in the benchmarks have been adjusted. This will be noted separately.


This time we decided on a somewhat broader range of tests. On the one hand, to get a more representative overview of the performance spectrum of the cards and, on the other hand, to counteract the special driver optimizations of the chip manufacturers for the two most popular benchmarks, Quake3 and 3DMark2001, which have recently become more and more popular.

Synthetic benchmarks
  • 3DMark 2000 v1.1
  • 3DMark2001 SE
  • Villagemark D3D v1.19
  • GL Excess v1.1a
  • VulpineGL v1.1p
Game benchmarks
  • Aquanox v1.17
  • Aquamark v2.3
  • DronezMark
  • Giants v1.4
  • Unreal Tournament v4.36
  • Quake III Arena v1.30
  • Serious Sam v1.02
  • Max Payne v1.02
FSAA and overclocking benchmarks
  • Quake III Arena v1.30
  • 3DMark2001 SE
  • Giants v1. 4

On the next page: 3DMark 2000 v1.1