RTSS generally offers better performance and more features compared to NVCP v3 Limiter.
RTSS generally offers better performance and more features compared to NVCP v3 Limiter.
The situation has evolved since the video was posted. Recent updates and driver releases have likely improved frame rate limiters, bringing input lag closer to in-game FPS limits. Frame times have seen adjustments, and MSI Afterburner’s OSD offers valuable insights for monitoring. Based on testing, one tool may provide more consistent and stable frame times than the other. Let’s compare them to find the best option for minimizing lag.
Still not great, but at least the frame times and software matched up in-game. I mainly use NVIDIA settings like frame limiter with ULLM, G-Sync, and a 141fps cap. I also apply Reflex with its boost options when it works. It helps that OW and FN have been updated, though they’ve made a lot of the settings harder to read.
Thanks for the update! I also enjoy Overwatch, so appreciate you sharing your configuration. Besides OW, I play other competitive titles such as CS GO, Starcraft 2, fighting games, and DotA 2. Most of these don’t heavily rely on the GPU, actually only a few do. For instance: Starcraft 2 LotV and many fighting games are capped at 60 FPS by default. My GPU usage stays around 22-50% (depending on the game), while CPU usage changes across different titles and cores—some games use just one, two, or all four cores.
While testing with RTSS and NVCP MFR, I noticed that without any restrictions, frame times stayed around 16.6 ms (with a slight variation). When either limiter was applied at 60 FPS, frame times settled quickly and stayed steady. The OSD graph showed a smooth line without spikes. Using NVCP’s MFR caused frame times to stabilize at 16.6 ms, with no fluctuations. CPU usage stayed consistent, but GPU usage increased when using NVCP MFR. When I ran RTSS, GPU usage didn’t change much, but NVCP MFR caused a noticeable jump in GPU load.
With this in mind, is it normal for MFR to behave this way? Or could there be an issue with the driver? Most of these tests were tracked using MSI Afterburner’s OSD.
Also, you mentioned using ULLM. Doesn’t Low Latency Mode affect latency only when your game is very GPU-intensive (using 97-99% of the GPU)?
This feature appears to adjust the number of frames in the render queue before sending them to the GPU. It uses "On" and "Ultra" as shorthand for MPRF values of 1 and 0. It's not the same as NVIDIA Reflex, which is a different tool for improving developer performance. Its impact on FPS in games depends on the game type—especially if they rely heavily on GPU or CPU resources. It might not necessarily lower input latency and could even increase it in some cases. More details can be found in the videos linked above.