I currently have a GTX 980 graphics card that runs Zwift at a decent frame rate (40fps+) rendered at 4K, although downsampled to my 1200 line monitor. The graphics look great but I’m always looking out for better.
My question is: If I was to buy the forthcoming GTX 1080 card, would the game utilise the extra power by increasing the number of polygons per frame or would that be fixed and I’d simply get a higher frame count?
I’m not sure whether this metric is fixed by Zwift developers to run on the lowest common denominator setup or whether it dynamically changes to a more detailed model if the graphics card (+cpu) is capable.
Internally Zwift categorizes cards into 4 types, basic, medium, high, and ultra. Those 4 settings determine polygon counts and special FX presets.
A 980 is already set to the highest settings automatically, so what you’d get by buying the 1080 is the ability to run at 4k at a much higher framerate.
Thanks Jon, that’s good to know. Does the Ultra polygon count match the limit of what the artists use to model everything or are there more polygons waiting to be ‘unlocked’ by future generations of graphics cards or in game settings to trade frame rate for details for example? eg: If I’m happy with 40fps at 4K and buy a card with twice the performance, could I unlock twice the detail @ 40fps via a config file option?
Apologies if this is a basic question. I know how chips work as I design them, but I’m not sure how the 3D artists model their world.
Also, do the Zwift settings overwrite any Nvidia specific settings for the game that I might set? eg: anti-alias oversampling.
Finally, the graphics in their current state are really superb and now beat the Tacx VR offering, which was my past current comparison point. What would you say is the frame rate limit that offers diminishing returns beyond and would I notice much difference in actually buying a 4K monitor. I’m aware that the pixel density is approaching the limit of what the eye can discern depending on the size and distance sat from it.