For example, in Overwatch that game has its own internal FPS limiter which has been shown by channel's such as BattleNonSense to be a "proper" in-engine FPS limiter which reduces input latency VS something like an external limiter such as RTSS or Nvidia's driver level limiter. They also had an internal setting called "Reduce Buffering" before "Nvidia Reflex" was added to the game which is still present there. BattleNonSense did put out a new video which seemed to show that Reflex is the better option for reducing latency, but there seems to one caveat which would be quite meaningful which I don't recall him addressing which has to do with G-Sync and Nvidia enforcing an FPS limit in conjunction with ULLM mode or Reflex. So, I have a 144 Hz panel and in Overwatch I cap the FPS in-engine/internally to 140 for lowest latency (as opposed to using an external limiter like Nvidia's). Now, if I turn Nvidia Reflex mode to ON in the game's settings this caps the game to 138 fps internally (to avoid hitting the g-sync ceiling presumably). The thing is though, I do NOT want to use the Nvidia driver level limiter -- I want to use the game's own internal limiter, but I'm not certain which its using. I would presume its using the Driver level limiter as that's what happens with the ULLM mode + G-Sync and Overwatch ignores my own manually placed setting at 140 FPS limit. Does Nvidia Reflex + G-Sync use the game's own limiter or Nvidia's driver level one? My bet would be its their own driver limiter one, but there should be some way to manually disable this automatically enforced limit because the user might prefer to cap the game's framerate either with a game's own internal limiter or with RTSS. You can kind of get around this by capping with the game's internal limiter to something like ~3 frames beneath where Nvidia's is capping, but then you're just wasting overhead and it seems like bad practice to have two fps limiters ON for a game simultaneously. I had made a thread on this in the past here: https://forums.guru3d.com/threads/a...-to-remove-the-auto-fps-cap-from-ullm.432794/ And I was informed that this behavior is "intended" and unlikely to change, but frankly that is stupid to be blunt. If a player's aim is competitive E-Sports and they want to achieve the lowest latency possible then there needs to be a way to disable this auto-fps limit so that users can use the lower latency further with in-engine limiters. I would personally also rather use RTSS over the Driver level limiter, but Nvidia doesn't let me choose. One solution would simply be to respect the user's manually set driver level FPS limit in the control panel -- so, if the game would ordinarily cap the FPS to 138 on a 144 Hz G-Sync panel "automatically" with the driver level limiter when ULLM or Reflex is ON, if the user manually goes into that game's profile and sets the limit either to a different value (like 140 instead of 138) or if a user turns the limiter off it should be respected. As it is, if you're reaching the G-Sync ceiling, you're probably better off with Reflex OFF then just using the in-game limiter set to "Refresh -3" or "Refresh-4" as per the blurbuster's tests and perhaps Reduce Buffering ON.