![metro last light benchmark vs actual game metro last light benchmark vs actual game](https://cdn.wccftech.com/wp-content/uploads/2020/10/abhjs016.jpg)
Sure, it'll only help in certain games, but the same could be said for a CPU upgrade, yet that doesn't seem to stop most people.Īs my results show, a PhysX card can be of great value to a high-end rig. Even if you don't, it may well be worth it buying a new card specially for PhysX: this may well be one of the more economical upgrades available, as you can get a significant difference from a comparatively cheap piece of hardware. If you've got an older card left over from an upgrade, it's a no-brainer: keep it for PhysX. Honestly, I think almost anyone with an Nvidia GPU would benefit.
![metro last light benchmark vs actual game metro last light benchmark vs actual game](https://images.gnwcdn.com/2013/articles/1/5/8/0/9/1/1/136877974342.jpg)
Let's see how our dedicated GTX 650 holds up. O rigin stakes these a notch up, with dynamic snow, steam and fog that reacts to the player. The Batman: Arkham games have always made impressive use of PhysX, using them not just for splashy tricks, but to subtly enrich the general atmosphere of the gameworld. Still, I'd call this a win for the PhysX card, given the significant average fps improvement, and thinner line overall. But, unfortunately, it also adds more occasional spikes and dips. On the whole, adding a PhysX card significantly thins the black line out, improving framerate consistency.
![metro last light benchmark vs actual game metro last light benchmark vs actual game](https://cdn.shortpixel.ai/spai/q_lossy,ret_wait,to_webp/https://www.gamingpcbuilder.com/wp-content/uploads/fov-68-min.jpg)
The results are more of a mixed bag in single-GPU mode, which already has a much thinner line (less variability) than the SLI system. Though still not perfect, the stutters were much less frequent, especially in the latter, more PhysX-intensive half of the benchmark.
#Metro last light benchmark vs actual game full
Without a PhysX card, the benchmark was so full of microstuttering I found it somewhat frustrating to watch. I can confirm that this was very noticeable when watching the actual benchmarks. Though the overall FPS is actually pretty similar, it's a much smoother ride. With an added PhysX card, the graph has the same general shape - the same peaks and troughs - but with much less variation. Without a PhysX card, the line jumps wildly all over the place, routinely making leaps of 50 fps or more in a single split second. That's the downside of multi-GPU setups, unfortunately, and it seems particularly pronounced in Metro Last Light.īut beyond that, notice the large difference in the black line of both graphs once you introduce a PhysX card, especially in SLI. The first thing you'll notice is the much thicker black lines on the SLI systems, revealing a much greater rate of framerate variability (ie.