Game Testing
Now that we have an idea of the performance gains you might see between the two generations of systems for applications we were curious about a games and graphics. Since the GTX 570 is basically an updated FERMI version of the GTX 470, we did not expect to see significant gains in performance. The GTX 470 was a Super Clocked EVGA version and the GTX 570 is the Classified version from EVGA. We ran all these tests at both the maximum resolution of our 30″ display, 2560 x 1600 and a more common 1920 x 1200.
Aliens vs Predator D3D11 Benchmark v1.03
This is a stand alone benchmark based on the Aliens vs. Predator game we use frequently around Legit Reviews to test our graphics cards with DirectX 11 using heavy tessellation.
Our guess about not being a big difference was fairly accurate with this test showing modest 15% and 14% improvement between the systems. The refinement in the GF110 in conjunction with the updated motherboard chip set and newest CPU is showing the advantages of current generation gear.
S.T.A.L.K.E.R.: Call of Pripyat
S.T.A.L.K.E.R CoP is another standard benchmark we have been running lately to push the DirectX 11 load on our GPU’s. This game / benchmark uses XRAY 1.6 Engine and a lot of tessellation and Enhanced Full Dynamic Lighting.
This benchmark from the S.T.A.L.K.E.R CoP game showed a 16% and 14% improvement moving to Sandy Bridge / GF110 system. At the highest resolution this is a 7 FPS gain which could make the difference between playable and laggy in some games.
Metro 2033
Metro 2033 is an extremely demanding DirectX 11 game that has made many a GPU cry mercy. This game utilizes the multi-threaded 4A Engine to really taxes these GPU’s. The engine usesvolumetric fog, film grain effects, sub-surface scattering, HDR tone mapping, object blur, and parallax mapping. The settings were MSAA 4x, AF16x, and all other settings set to maximum.
While both GPU’s struggled a bit at the 2560 x 1600 resolution giving us OK but not really smooth playable gaming the new system actually wasn’t bad. Running at 25.2 FPS the new rig was getting close enough to the ideal minimum of 30 FPS that just dialing down some of the eye candy a notch should put this title above that mark for a fully playable experience. The 17% and 19% improvement is again consistent with the other gaming benchmarks.
Conclusion
While we did not see the 35% improvement that 3D Mark Vantage showed we did see an average of a 16% improvement in graphics performance across all these titles. To be fair these are DirectX 11 titles and as we said earlier 3D Mark Vantage is a DirectX 10 engine which could be the reason for the difference in percentages. For me a 16% average improvement is about when I start thinking it could be worth it to jump to the next generation hardware. Over 25% and I feel like I am way behind the curve.
Comments are closed.