A GeForce PC can expose two settings that sound as if they do the same job. A supported game may have NVIDIA Reflex Low Latency in its own graphics menu, while NVIDIA Control Panel has a driver-level Low Latency Mode with an Ultra option. Turning on every low-latency switch feels logical, but the technologies operate at different layers.
NVIDIA’s own guidance is straightforward: when a game supports Reflex, use Reflex. The driver option is mainly the fallback for games that do not have a Reflex integration.
Why render queues add latency
A CPU can prepare rendering work faster than the GPU can finish it. When work accumulates in a render queue, a frame can represent input that was sampled earlier than necessary. You may still see a high FPS number while the click-to-photon response is less immediate.
Low-latency technologies try to prevent excessive work from waiting in that queue. The key difference is how much of the game pipeline each method can control.
What the driver Ultra Low Latency Mode does
NVIDIA’s driver-level Ultra Low Latency Mode reduces the render queue from outside the game. It can be useful when the title does not implement Reflex because the driver can still change how rendering work is queued.
But the driver does not own the entire game simulation and submission pipeline. NVIDIA describes this as a partial latency optimization compared with a native Reflex integration.
Why Reflex can do more
Reflex is integrated into the game engine. NVIDIA says Reflex Low Latency synchronizes CPU and GPU work so rendering is submitted just in time, which can remove the render queue and reduce CPU back-pressure in GPU-bound scenarios.
Because the game participates directly, Reflex can control timing earlier in the pipeline than a driver-only switch. That is the reason NVIDIA recommends Reflex over Ultra Low Latency Mode when both are available.
What happens if you enable both?
You do not need to worry that the two features will stack into some unstable double optimization. NVIDIA states that Reflex operates independently of the driver Ultra Low Latency setting and that Reflex takes priority when both are enabled.
For a clean configuration, I would still avoid unnecessary settings. If the game has Reflex, enable it in the game and leave the driver Low Latency Mode at its normal setting unless you have a specific reason to test otherwise.
On versus On + Boost
Some games expose Reflex as On and On + Boost. NVIDIA explains that the Boost behavior keeps GPU clocks higher in situations where the GPU is underutilized, which can shave additional latency in some CPU-bound cases. The tradeoff is higher power consumption.
If you are on a desktop and chasing minimum competitive latency, Boost is a reasonable test. On a laptop, the extra power and heat can matter more. Compare the actual latency or frame-time result rather than assuming the more aggressive label is always superior.
Higher FPS and lower latency are related but not identical
FPS measures throughput: how many frames are produced per second. System latency measures how long it takes an input to be reflected on screen. Raising FPS often reduces latency because frames are produced more frequently, but a render queue can still add delay.
This is why a latency feature can make a game feel more responsive without producing a large FPS increase. Reflex is designed to change pipeline timing, not to act as an upscaler or add rendering performance.
Do not destroy image quality just to keep GPU usage low
Older latency advice often told players to turn every graphics setting to minimum so the GPU was never busy. NVIDIA’s Reflex material specifically targets GPU-bound latency by preventing work from piling up, which gives players more room to choose graphics settings without the same queue penalty.
There is still a limit: a slower rendered frame takes longer than a fast one. But you can tune image quality, FPS and latency as separate goals instead of assuming “low settings” is the only correct competitive configuration.
A simple configuration order
First, check whether the game has a Reflex option. If it does, enable Reflex and test On versus On + Boost if available. If Reflex is absent, the NVIDIA driver Ultra Low Latency setting becomes the more relevant fallback. Then verify that your monitor is actually running at its maximum intended refresh rate.
Test with the same game scene and avoid changing V-Sync, frame caps, graphics quality and low-latency mode all at once. Latency tuning is much easier when you know which change produced the result.
Which setting should you use?
If Reflex is supported, use Reflex. That is NVIDIA’s documented preference because the game-integrated path can manage the pipeline more effectively than the driver alone. If Reflex is not supported, Ultra Low Latency Mode is a useful driver-level option to test.
Do not treat the two switches as FPS boosters. Their job is responsiveness. The best result is the configuration that keeps frame delivery stable while reducing the time between your input and the visible response.
