Terminology:
-HAGS= Hardware Accelerated GPU Scheduler
-A GPU bound situation is when the GPU reaches 95+%, Doing so means the CPU draws frames into a buffer for the GPU to render later and there is when latency is created.
-GPU limit is the time period of which the GPU is bound.
Currently, Enabling HAGS with Nvidia reflex without a frame cap can cause worse latency than disabled HAGS with Reflex.
From what I know the additional GPU workload of HAGS adds to reflex which cause minor amounts of GPU limit time.
In Overwatch there are drastic frame drops with HAGS and reflex enabled, but that could be the stress testing methods used.
If you are aiming for an uncapped FPS target the Nvidia reflex can do a reasonable job of keeping you out of a GPU bound situation without HAGS.
If you have a FPS target in mind then using HAGS can improve framerates if your GPU can remain in a unbound scenario.
Additionally, disabling Nvidia Reflex will give you more FPS. (Because Reflex is designed to avoid GPU bound scenarios a frame cap does the same thing only that its not a dynamic framerate.)
Here are some results from Apex: (uncapped)
The yellow and orange lines have higher framrates but they are GPU bound, this is where using a good framerate limiter like RTSS comes in to find the sweet spot as in game frame rate limiters are erratic.
So anyone here using Low Latency V-sync or just capping to a specific target will benefit a lot from HAGS.
HAGS was introduced in 2020 so its rather new-ish.
Devblog:
https://devblogs.microsoft.com/directx/ ... cheduling/