[kimixa]

I find I can't really tell the difference beyond ~90fps outside of artificial tests (like moving a hard edge and "counting" the edges visible, you can normally do this with a mouse cursor) in non-interactive content.

But beyond that the latency improvement is still noticeable - most games still tend to handle input and game update in terms of frames - so if you have 3-4 frames of input->visible result latency that's 30-40ms at 100fps, but 15-20ms at 200, and you can tell the difference in some situations. This is very dependent on the style of game though - many have sluggish movement and slow direction changes as part of the design, so is less noticeable than some super slide-y shooter like Doom.

But much of that is solvable by the app outside of higher FPS, with GPU submission timing and delaying input processing you can get pretty much the same latency at any FPS, just many games don't do that (yet?)

[trafficante]

Interesting point about gaming input latency: I haven’t been really interested in twitchy flatscreen FPS since the early days of Overwatch 1 and my VR stuff is mostly bottlenecked by network and decode latency.