Synthetic Benchmarks

As usual we start of the benchmark analysis by going through the different benchmark suites coming from Futuremark, and the rather new RyderMark program. These will provide a good indication on the game play of today's games. First we will start with RyderMark, followed by the different gaming benchmark programs from Futuremark such as 3DMark03, 3DMark05 and last but not least 3DMark06. We will reboot our system each time for each of the four synthetic API's used and leave all processes running in the background. We put the Jetway Radeon HD 2900 XT accelerator against the nearest competition from NVIDIA, the GeForce 8800 GTS card provided by Gainward, and analyze the performance figures. So let's just give it a go...

Candella RyderMark

Just recently a new benchmark suite came to the market, developed by Candella Software. The benchmark called RyderMark utilizes a proprietary multithreaded game engine developed for a game scheduled for future release, and does require a high-end personal computer featuring a dual core microprocessor as well as DirectX 9.0c, Shader Model 3.0 capable graphics card with 512MB of memory onboard. This benchmark suite uses true 64-bit High Dynamic Range lighting with anti-aliasing, parallax occlusion mapping, soft shadows, normal mapping, soft particles, full scene motion blur, depth-of-field, heat haze, volumetric fire and realistic water physics.

Benchmark Results - 1280x1024 - E6300 2.80GHz - 2GB RAM
	GeForce 8800 GTS 640MB	RyderMark Score
SM2.0/No AA	513/1584MHz (Ref - Single)	3460
SM3.0/No AA	513/1584MHz (Ref - Single)	3435
SM3.0/4x AA	513/1584MHz (Ref - Single)	2904
	Radeon HD 2900 XT 512MB	RyderMark Score
SM2.0/No AA	742/1658MHz (Ref - Single)	3213
SM3.0/No AA	742/1658MHz (Ref - Single)	3132
SM3.0/4x AA	742/1658MHz (Ref - Single)	2044
Results Interpretation: Higher index values are better.
Benchmark Results - 1600x1200 - E6300 2.80GHz - 2GB RAM
	GeForce 8800 GTS 640MB	RyderMark Score
0x AA/Optimal	513/1584MHz (Ref - Single)	3107
4x AA/8x AF	513/1584MHz (Ref - Single)	3102
8x AA/16x AF	513/1584MHz (Ref - Single)	2568
	Radeon HD 2900 XT 512MB	RyderMark Score
0x AA/Optimal	742/1658MHz (Ref - Single)	2978
4x AA/8x AF	742/1658MHz (Ref - Single)	2781
8x AA/16x AF	742/1658MHz (Ref - Single)	1613
Results Interpretation: Higher index values are better.

It is a first for everything and we used for the first time the RyderMark benchmark engine. To compare our reported results we are using the GeForce 8800 GTS, a reference NVIDIA design graphics card. Our config build around the Jetway Radeon HD 2900 XT clocks down some outstanding scores, but when looking further to the GeForce 8800 GTS edition just leaves it way behind in terms of raw performance. The figures show that enabling anti-aliasing is in favor of the crippled G80 chip, and doesn't drop that fast in performance. Since RyderMark is rather new we don't have yet a good view on what is influencing the score most, although when using the same system it shows that the GeForce 8800 GTS is in favor. The Radeon HD 2900 XT is not far behind, but is not able to keep up.

Benchmark Results - 2560x1600 - E6300 2.80GHz - 2GB RAM
	GeForce 8800 GTS 640MB	RyderMark Score
0x AA/Optimal	513/1584MHz (Ref - Single)	2430
4x AA/8x AF	513/1584MHz (Ref - Single)	2327
8x AA/16x AF	513/1584MHz (Ref - Single)	2122
	Radeon HD 2900 XT 512MB	RyderMark Score
0x AA/Optimal	742/1658MHz (Ref - Single)	2543
4x AA/8x AF	742/1658MHz (Ref - Single)	2462
8x AA/16x AF	742/1658MHz (Ref - Single)	1249
Results Interpretation: Higher index values are better.

We are really into the HD gaming, which means we did have to run the benchmark at the highest widescreen resolution possible on our UltraSharp 30" LCD display. This somehow turned the tables slightly in favor of the Radeon HD 2900 XT graphics card, which is due to the higher core clock speed. With the core set at 742MHz which is about 230MHz more than for the GeForce 8800 GTS it doesn't come as a surprise that with greater textures and more complex rendering of the scenes all comes down on how fast the GPU is clocked. And clearly this is in favor of the Radeon HD 2900 XT accelerator. But once we enable the anti-aliasing further the GeForce 8800 GTS does come out on top with a rather crushing score.

Futuremark 3DMark03

To continue we first go back to actually a quite older benchmark suite from Futuremark. By combining full DirectX 9.0a support with completely new tests and graphics, 3DMark03 continues the legacy of being the industry standard benchmark. The high quality game tests, image quality tests, sound tests and others give you an extremely accurate overview of your system's current gaming performance. So the synthetic analysis begins by taking a spin with 3DMark03 which is build around the DirectX 9.0 specifications, although still using quite some DirectX 8.1 features.

Benchmark Results - 1280x1024 - E6300 2.80GHz - 2GB RAM
	GeForce 8800 GTS 640MB	3DMark03 Score
0x AA/Optimal	513/1584MHz (Ref - Single)	24315
4x AA/8x AF	513/1584MHz (Ref - Single)	14653
8x AA/16x AF	513/1584MHz (Ref - Single)	10299
	Radeon HD 2900 XT 512MB	3DMark03 Score
0x AA/Optimal	742/1658MHz (Ref - Single)	28657
4x AA/8x AF	742/1658MHz (Ref - Single)	12029
8x AA/16x AF	742/1658MHz (Ref - Single)	8018
Results Interpretation: Higher index values are better.
Benchmark Results - 1600x1200 - E6300 2.80GHz - 2GB RAM
	GeForce 8800 GTS 640MB	3DMark03 Score
0x AA/Optimal	513/1584MHz (Ref - Single)	20146
4x AA/8x AF	513/1584MHz (Ref - Single)	11583
8x AA/16x AF	513/1584MHz (Ref - Single)	7361
	Radeon HD 2900 XT 512MB	3DMark03 Score
0x AA/Optimal	742/1658MHz (Ref - Single)	22861
4x AA/8x AF	742/1658MHz (Ref - Single)	9499
8x AA/16x AF	742/1658MHz (Ref - Single)	6225
Results Interpretation: Higher index values are better.

Being an already older benchmark, it does still perform quite well to show a first indication on the performance of the graphics accelerator. When looking at the results from the reference Radeon HD 2900 XT we can say that it is fast and does a bit better than the GeForce 8800 GTS. However this is changing a bit when we enabled the fancy features such as anti-aliasing and anisotropic filtering, as the GeForce 8800 GTS comes out on top. The Radeon HD 2900 XT can somehow keep up the pace, although can't keep up. The new architecture of unified shaders does well even with the older gaming engine, and you can as well enjoy the new visual effects of DirectX 10 in the near future.

Benchmark Results - 2560x1600 - E6300 2.80GHz - 2GB RAM
	GeForce 8800 GTS 640MB	3DMark03 Score
0x AA/Optimal	513/1584MHz (Ref - Single)	11942
4x AA/8x AF	513/1584MHz (Ref - Single)	5689
8x AA/16x AF	513/1584MHz (Ref - Single)	3062
	Radeon HD 2900 XT 512MB	3DMark03 Score
0x AA/Optimal	742/1658MHz (Ref - Single)	12786
4x AA/8x AF	742/1658Hz (Ref - Single)	5325
8x AA/16x AF	742/1658MHz (Ref - Single)	3503
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 ATI has jumped on the High Definition gaming wagon and show that by the new naming convention with the HD abbreviation in their product naming. Again we put both our setups against each other, with once again the Jetway config taking the lead. In the end it doesn't come that much as a surprise as 3DMark03 is not a full blown DirectX 9.0c benchmark suite, but rather limited to the lower versions. And to that extend the Radeon HD 2900 XT version just has better specifications and raw power to get the job done.