Menu
Asus V8440 TD and V8460 ultra TD in the test: Two 'Ti' tans among themselves

Asus V8440 TD and V8460 ultra TD in the test: Two 'Ti' tans among themselves

Templemark D3D

The Templemark, just like the following Villagemark, was originally a demonstration program by PowerVR. However, since it supports a lot of current features such as hardware TnL, bump mapping and up to 6 texture layers in one pass, it is also well suited as an independent oneBenchmark that is guaranteed not to be optimized for nVidia chips.

TempleMark D3D v1.06
  • 1024x768x32:
    • Asus V8460 ultra
      132.8
    • Asus V8440
      123,6
    • GF4 Ti4200 (250/222MHz)
      109.2
    • GF4 Ti4200 (225/250MHz)
      106.4
    • GeForce3
      88.4
    • GeForce3 Ti200
      77.9
    • GeForce4 MX440
      57.0

Unfortunately, the benchmark does not run in other resolutions, so in this case a single value must be sufficient. Without large numbers of polygons, the memory controller and the generally available bandwidth as well as the removal of invisible surfaces (HSR Hidden Surface Removal) mainly affect the frame rate achieved. However, this depends on the chip clock, and this is how it can be explained that in the back seats, the Ti4200 with the lower memory clock, i.e. the 128MB model, our 'prototype' version of the Ti4200 with low chip and high memory clock is the red one Lantern.

For this test, which is apparently very well suited to assessing the efficiency of the memory subsystem and the HSR, we once ran the GeForce4 at the same rate as the GeForce3 to see the improvements of the LMA-II net.

TempleMark D3D v1.06
  • Efficiency comparison:
    • GF4 Ti @ 200/230MHz
      95.1
    • GeForce3
      88,4

7 frames per second can only be gained in a texture-intensive scenario by optimizing the LMA-II compared to its predecessor in the GeForce3. In addition, the GeForce4 Ti cards usually have a significantly higher clock rate. So overall a very positive picture, as the higher clock rate is converted very well into higher performance.

Villagemark D3D

This benchmark was developed by PowerVR and is used to demonstrate the advantages of the Kyro2 to make clear that in VillageMark the overdraw is particularly large with a factor of up to 10. Most graphics cards also calculate the objects that are hidden by others and therefore actually only mean wasted bandwidth.

Villagemark D3D v1.19
  • 1024x768:
    • Asus V8460 ultra
      96
    • Asus V8440
      87
    • GF4 Ti4200 (250/222MHz)
      78
    • GF4 Ti4200 (225/250MHz)
      72
    • GeForce3
      63
    • GeForce3 Ti200
      55
    • GeForce4 MX440
      45
  • 1152x864:
    • Asus V8460 ultra
      79
    • Asus V8440
      71
    • GF4 Ti4200 (250/222MHz)
      63
    • GF4 Ti4200 (225/250MHz)
      59
    • GeForce3
      52
    • GeForce3 Ti200
      46
    • GeForce4 MX440
      37
  • 1280x1024:
    • Asus V8460 ultra
      61
    • Asus V8440
      56
    • GF4 Ti4200 (250/222MHz )
      49
    • GF4 Ti4200 (225/250MHz)
      46
    • GeForce3
      42
    • GeForce3 Ti200
      37
    • GeForce4 MX440
      28
  • 1600x1200:
    • Asus V8460 ultra
      44
    • Asus V8440
      40
    • GF4 Ti4200 (250/222MHz)
      35
    • GF4 Ti4200 (225/250MHz)
      33
    • GeForce3
      30
    • GeForce3 Ti200
      26
    • GeForce4 MX440
      20

Furthermore, is veryInteresting that the GeForce4 Ti doesn't seem to depend too much on the memory speed when a high overdraw comes into play. The bottleneck here is clearly in the chip and its ability to perform HSR sufficiently quickly. So you wouldn't be wrong in this scenario with a Ti4200 with 128MB lower clocked RAM.

On the next page: GL Excess

Comments