NVIDIA GeForce FX 5800 Ultra: It's Here, but is it Good?

AA+AF Performance - Codecreatures CodeBench

Codecult's Codecreatures engine has been turned into a fairly interesting DirectX 8.1 benchmark. Without widespread use of the engine this test ends up being no better than a synthetic benchmark, but it's interesting to see the results to help get an idea for how these cards perform under intense DirectX 8.1 applications.

Codecult Codecreatures AA/Aniso Performance 1024x768x32 GeForce FX 5800 Ultra (4X AA/8X Bal Aniso) GeForce FX 5800 (4X AA/8X Bal Ansio) Radeon 9700 Pro (4X AA/8X Perf Aniso) Radeon 9700 (4X AA/8X Perf Aniso) Radeon 9700 Pro (4X AA/8X Qual Aniso) Radeon 9700 (4X AA/8X Qual Aniso) 30.8 26.4 24.4 21 20.2 17.4 | 0 | 6 | 12 | 18 | 25 | 31 | 37

What's interesting here is that the Codecreatures benchmark is the first situation where the GeForce FX 5800 Ultra actually exhibits a clear performance lead, in this case a 26% performance advantage over the Radeon 9700 Pro. Could this be a hint of the GeForce FX's potential? Unfortunately it's entirely too early to say for sure, it could just be that this benchmark is stressing the higher performing pixel shader engines of the GeForce FX.

Codecult Codecreatures AA/Aniso Performance 1280x1024x32 GeForce FX 5800 Ultra (4X AA/8X Bal Aniso) GeForce FX 5800 (4X AA/8X Bal Ansio) Radeon 9700 Pro (4X AA/8X Perf Aniso) Radeon 9700 (4X AA/8X Perf Aniso) Radeon 9700 Pro (4X AA/8X Qual Aniso) Radeon 9700 (4X AA/8X Qual Aniso) 18.3 16.2 14.6 13.2 10.7 9.9 | 0 | 4 | 7 | 11 | 15 | 18 | 22

Nothing new at 1280x1024, the GeForce FX 5800 Ultra hangs on to a 25% performance advantage.

Codecult Codecreatures AA/Aniso Performance 1600x1200x32 GeForce FX 5800 Ultra (4X AA/8X Bal Aniso) GeForce FX 5800 (4X AA/8X Bal Ansio) Radeon 9700 Pro (4X AA/8X Perf Aniso) Radeon 9700 (4X AA/8X Perf Aniso) Radeon 9700 Pro (4X AA/8X Qual Aniso) Radeon 9700 (4X AA/8X Qual Aniso) 14.8 12.9 10.5 9.7 7.3 6.9 | 0 | 3 | 6 | 9 | 12 | 15 | 18

And we close things out with a 40% advantage at 1600x1200. It is too bad that this is only a synthetic test and we haven't been able to see even remotely similar results in any of our real-world benchmarks.