I think cloud gaming only takes off when the games are built specifically for this use case.
Right now, streaming gaming consists of encoding GPU output, relaying player inputs, and praying everything in between doesn't notice the player isn't in the same room anymore.
Think about how much compute and memory resources are wasted because we are running thousands of nearly duplicate gaming environments with wildly inconsistent utilization throughout. How much latency and resource consumption could be shaved off if the game was aware of its circumstances and engineered for this purpose?
I’ve generally had a good experience with Xbox cloud gaming. I’ve been playing city skylines on xcloud and decided to try downloading it to my xbox one s since my Bay Area internet keeps going it. Playing the game locally was super laggy and I only lasted a few minutes before deleting the game and playing via xcloud again.
You couldn't save on memory if everyone was using the same textures?
The most hacky but also potentially elegant approach would be to create/modify a game engine such that it can render 4 clients at a time onto the same viewport (i.e. 2x2 1080/720p). Just like the good old days with goldeneye64 multiplayer. With a setup like this, you could take advantage of cost sharing throughout.
Again, would have to build the game from zero to target this path. I think there could be some economic arguments to be made here. The software and creativity are the only real limitations.
Right now, streaming gaming consists of encoding GPU output, relaying player inputs, and praying everything in between doesn't notice the player isn't in the same room anymore.
Think about how much compute and memory resources are wasted because we are running thousands of nearly duplicate gaming environments with wildly inconsistent utilization throughout. How much latency and resource consumption could be shaved off if the game was aware of its circumstances and engineered for this purpose?