|
Page 3 of 6 

Synthetic Benchmarks (Cont.) We remain still with the synthetic benchmarking of the Point of View GeForce 8600 GTS graphics card and how it stacks up against the PowerColor Radeon X1950 PRO accelerator. We will go through the newest benchmark suites developed by Futuremark. These will provide a good indication on the processing value and performance of the GeForce 8800 GTS, but next to that also on the game play of today's games. Similar to the other benchmarks we will compare both configuration and analyze the differences related to the performance figures. So, read on and find out what it is all about...
Futuremark 3DMark05 
The following version from Futuremark came as 3DMark05, and is best suited for the latest generation of DirectX 9.0 graphics cards. It is the first benchmark to require a DirectX 9.0 compliant hardware with support for Pixel Shaders 2.0 or higher! By combining high quality 3D tests, CPU tests, feature tests, image quality tools, and much more, 3DMark05 is a premium benchmark for evaluating the latest generation of gaming hardware. In fact 3DMark05 is the answer to the continuously growing challenge in benchmarking. Designed for higher image quality and the new processing capabilities, it will show some input and what is possible with compliant hardware. | Benchmark Results - 1280x1024 - E6300 2.80GHz - 2GB RAM | | Radeon X1950 PRO 256MB | 3DMark05 Score | | 0x AA/Optimal | 580/1380MHz (Ref - Single) | 9289 | | 4x AA/16x AF | 580/1380MHz (Ref - Single) | 7425 | |
| GeForce 8600 GTS 256MB | 3DMark05 Score | | 0x AA/Optimal | 675/2000MHz (Ref - Single) | 9379 | | 4x AA/16x AF | 675/2000MHz (Ref - Single) | 6919 | | Results Interpretation: Higher index values are better. |
Once again we can see a similar result while using 3DMark05, a small win for the GeForce 8600 GTS with the default settings but not good enough to beat the Radeon X1950 PRO when extra features and image settings are enabled. The Radeon X1950 PRO has a higher memory bandwidth as it features a 256-bit memory interface which makes a difference when looking at frame buffering and shader intensive tasks. Which is as well a bit the bad thing about the GeForce 8600 GTS by only having a 128-bit memory interface and by doing so limiting the throughput from and to the memory of the graphics accelerator. | Benchmark Results - 2560x1600 - E6300 2.80GHz - 2GB RAM | | GeForce 8600 GTS 256MB | 3DMark05 Score | | 0x AA/Optimal | 675/2000MHz (Ref - Single) | 4700 | | 4x AA/16x AF | 675/2000MHz (Ref - Single) | 1860 | | Results Interpretation: Higher index values are better. |
You got that right, some HD gaming input as well as we are now using one of these lovely widescreen 30" LCD displays that feature an awesome resolution of 2560x1600 to experience just that extra more. And actually NVIDIA is quite a fan of these HD resolutions as they call it the Extreme High Definition experience. Of course with the GeForce 8600 GTS being a mid-range graphics card you expect a hit in the performance. And that is actually what you get. You can't expect that a mid-range card is able to deliver stunning performance at such a widescreen resolution, but the card does support it nonetheless. Which is a darn nice feature.
Futuremark 3DMark06 
To complete the view and Futuremark performance figures we also run a 3DMark06, which is the worldwide standard in advanced 3D game performance benchmarking. A fundamental tool for every company in the PC industry as well as PC users and gamers, 3DMark06 uses advanced real-time 3D game workloads to measure PC performance using a suite of DirectX 9.0c 3D graphics tests, CPU tests, and 3D feature tests. 3DMark06 tests include all new HDR/SM3.0 graphics tests, SM2.0 graphics tests, AI and physics driven single and multiple cores or processor CPU tests and a collection of comprehensive feature tests to reliably measure next generation gaming performance today. | Benchmark Results - 1280x1024 - E6300 2.80GHz - 2GB RAM | | Radeon X1950 PRO 256MB | 3DMark06 Score | | 0x AA/Optimal | 575/1380MHz (Ref - Single) | 5214 | | 4x AA/16x AF | 575/1380MHz (Ref - Single) | 4020 | |
| GeForce 8600 GTS 256MB | 3DMark06 Score | | 0x AA/Optimal | 675/2000MHz (Ref - Single) | 5647 | | 4x AA/16x AF | 675/2000MHz (Ref - Single) | 3951 | | Results Interpretation: Higher index values are better. |
Well, once again we don't notice much difference with the previous results although that the results are much closer together somehow linked to the unified shader implementation which seems to function better with DirectX 9.0c capable hardware. Nonetheless if NVIDIA would have foreseen the GeForce 8600 GTS with a 256-bit memory bus it might have been a different story. Certainly the memory interface is what it all comes down to when we go and enable the higher image quality options like antialiasing and anisotropic filtering... So, we could say that the GeForce 8600 GTS does well, but it might have done much more with a 256-bit memory interface. | Benchmark Results - 2560x1600 - E6300 2.80GHz - 2GB RAM | | GeForce 8600 GTS 256MB | 3DMark06 Score | | 0x AA/Optimal | 675/2000MHz (Ref - Single) | 2720 | | 4x AA/16x AF | 675/2000MHz (Ref - Single) | N/A | | Results Interpretation: Higher index values are better. |
Again we wanted to show as well the HD gaming experience that a GeForce 8600 GTS can deliver, and although the card supports the Extreme HD feature from NVIDIA it does only partly come to its full potential. The support is one thing, the performance and feeling is another. One thing to notice is that with only 256MB of frame buffer you won't be able to enabled antialiasing and/or anisotropic filtering to a high level. In the end that is somehow logic due to the large amount of textures to be rendered which will require at least 320MB to 512MB memory. If NVIDIA would like to fully support the Extreme HD experience a GeForce 8600 GTS with 512MB would be more than welcome. 
|