VSync can change latency, but there is no single penalty that applies to every Dynamic MFG setup. The result depends on how the installed integration schedules frames, whether it reaches the display ceiling and which pacing controls are active.
For the related variable-refresh setup, see configuring G-Sync with Dynamic MFG.
Do not rely on a permanent “Dynamic never supports VSync” rule. Support changed after launch. I would verify the actual game or override path before applying a synchronization preset.
Start with what VSync is trying to solve
VSync coordinates presentation with the display’s refresh process to address tearing. Waiting and backpressure can affect responsiveness, especially when the rendering and presentation path reaches a limiting boundary.
A variable-refresh display can change its refresh timing within its operating range. That is different from assuming all VSync behaviour disappears whenever VRR is enabled.
The useful question is not just whether a checkbox is on. It is whether the intended synchronization path is active, what rate it controls and how it behaves in the scene you play.
Separate a visible tear, an uneven pan and a delayed control response in your notes. One setting can change those symptoms differently.
Why early Dynamic advice is incomplete now
NVIDIA’s Streamline release history records VSync and frame-limiter support added to Dynamic mode in SDK 2.14.1 in September 2026. The earlier launch restriction should therefore not be repeated as a permanent capability limit.
That developer release does not update every shipped title automatically. The game, driver and override path still determine what is supported on your machine.
Current integration guidance also describes different target handling when VSync is active. A custom-target number need not govern the stream in the same way under every synchronization configuration.
Use official game and app updates. Do not assume that manually swapping a DLL is necessary or sufficient to adopt the complete newer integration.
The display ceiling matters
If the output approaches or exceeds a display boundary, the presentation path may behave differently from the same workload comfortably inside a supported VRR range. That is why a quiet scene can expose a different problem from a demanding scene.
Do not infer the active path merely because the counter appears a few frames below the monitor’s advertised refresh rate. Confirm VRR and the settings through supported indicators and documentation.
A ceiling is not a promise of performance. In a heavy scene below the target, a VSync comparison can be dominated by the workload instead of the ceiling behaviour.
I would test both an easy area that reaches the boundary and a demanding area that does not. Keep the route and other settings consistent between runs.
Avoid importing a generic preset unchanged
Advice for conventional rendering or fixed generation can be useful background, but Dynamic mode has its own integration requirements. Verify the installed combination before applying a GSync, VSync and external-cap recipe.
Identify whether VSync is controlled in-game or in the driver for your actual display mode and system. A setting that applies in one path may not govern another.
Start without extra manual caps unless the documented setup requires them. Several limits can make it difficult to tell which one changed the response.
The frame target guide explains why a sustainable output target and a base-rate limit are different decisions.
Run a comparison you can interpret
Record the game build, GPU, driver, refresh rate, window mode, Dynamic setting and all active caps. Save the existing configuration before testing.
Where both choices are supported, compare VSync on and off while holding quality, target and the route constant. Note tearing, pacing and control response separately.
If changing VSync also changes target handling, record that as part of the practical configuration difference. Do not pretend that the two runs isolate only one internal wait.
Return to the first setup and repeat. Without a defined latency measurement, report response feel as an observation rather than assigning a millisecond cost.
If VSync makes the controls feel worse
Confirm that the effect repeats in the same scene. Check for an unexpected base-rate drop, a competing limiter or behaviour at the display ceiling.
If the setup is not documented as supported, restore a known supported path before fine-tuning. An incompatible configuration is not a useful measurement of VSync’s normal cost.
The sister site’s RTSS limiter guide helps if an external cap is also active. Test the limit separately from the overlay instead of removing several tools at once.
If the supported tear-free path still feels too delayed for the game, compare another supported mode or accept tearing for a response-focused configuration. That is a preference with a visible trade-off, not a universal mandate.
My practical recommendation
Use the synchronization configuration supported by the installed Dynamic path and judge both tearing and response. Keep the tear-free option if its response trade-off is acceptable in the actual game.
If you cannot establish support or the result remains inconsistent, compare a documented fixed generation configuration. The goal is a clear, reliable setup rather than defending one checkbox in every situation.