So yesterday they announced some new technolgy tools for developers, allowing them to easily implement volumetric fire effects, physx based physics simulations of liquids and particles, and perhaps most importantly FULL GLOBAL ILLUMINATION on Nvidia GPU's.
Sweet mother of god.
They also mentioned some improvements coming with their Physx 3.x api such as full multi-threading.
Today they introduced Shadowplay (beta coming Oct 28th), which can buffer up to 20 minutes of video, streams directly to twitch, allows you to record as much video as you want at 1080p/60fps/up to 50 mbps, and supports video overlay of your webcam/other vidoe source. All for no FPS hit as the built in video encoder on Nvidia GPU's handles everything.
And finally, Nvidia has basically solved the issues with V-sync and screen tearing!
With G-Sync a 3 part solution to the problem, they will eliminate the need for Vsync (which introduces latency and stuttering), AND remove screen tearing that Vsync would normally solve. The 3 part solution is a Kepler GPU, your GPU drivers, and, the key part, a chip embedded into monitors that allows the GPU to drive monitor timing. They announced several partners including Asus who will be offering panels from 1080p all the way to 4K.
Carmack was also on hand, and I believe him and people from DICE are having a conference after the Nvidia show. So far he spoke about the huge difference in smoothness and image quality that G-sync brings to the table.
My guess is that Gsync is also goign to be critical for Oculus rift support. GPU driving the timing of the display will probably be of great help in terms of redicing latency - something Carmack has spoken of extensibly in the past as being the bane of virtual reality.
Log in to comment