hello what is this new NVIDIA reflex latency graphics option and what is it for
1 Like
Also, which are the differences (in DLSS mode) between AUTO, DLAA and QUALITY?
1 Like
Google is your friend: Introducing NVIDIA Reflex: Optimize and Measure Latency in Competitive Games | GeForce News | NVIDIA
1 Like
What is the difference between off, on and on + boost?
1 Like
To summarize that link:
- really long explanation of what latency is and how it affects FPS-style gameplay (for MSFS users, this would most affect people using VR or head-tracker display controls)
- an actual technical detail mentioned once, which is that it switches the GPU into high-frequency mode more of the time, so it doesn’t have to bump frequencies up and down and can maintain a smaller processing queue, thus reducing latency (nice! but – at the cost of increased power/heat)
3 Likes
thanks that you mentioned point 2) … most users often enable settings , but for which they at the end not see/notice any kind of “changes” nor that they had former any kind of issue. But lots of these “features” disable usually mainly most of energy-saveing-features on the gpu ( e.g. also that max perf mode ). It all like marketing gags or in best case noticable from ESL full-time gamers 
1 Like
Auto i guess will pick the (best)option for your monitor, don’t use it.
DLAA is DLSS but instead of aiming at better fps it brings better quality, cause the frame they use to kick their AI stuff in is in the same resolution as your monitor.( dlss is a lower resolution)
1 Like
Boost overrides the Power Management settings in the NCP and sets Power Management to maximum performance so the GPU doesn’t idle while the CPU is limited.
So, Low Latency Mode = Ultra plus Power Management = Prefer Max Performance in the NCP is about the same as Reflex + Boost in game settings.
3 Likes
but be warned that it do exactly that " it not clock down " , similar to Reflex… No more energy saving, no change to cooldown, etc… Max Perf is as a reason NOT the default setting
2 Likes
Yes, I am skeptical about boost. If the CPU frame time is 10ms and the GPU frametime is 7ms, I find it hard to believe the GPU would idle for 3ms waiting for the next frame from the CPU.
When I enable the Reflex option my CPU goes to 20ms or more and if I select the boost option my CPU jumps to over 30ms. No effect whatsoever on the GPU. With reflex off I have very good performance. Very odd. 5800x3d and 3080 12gb
none of this jives with what it does though…
To my understanding, this only works on RTX 40 Series GPUs, right? What happens if the setting is enabled with older GPUs - absolutely nothing or can it cause issues?
It applies to GTX 900 and later.