3 reasons DLSS 3’s frame generation isn’t everything it’s cracked up to be

Software features like Deep Learning Super Sampling (DLSS) have quickly become a staple in PC gaming, offering essentially free performance boosts. DLSS can upscale your game resolution in a way that’s often barely noticeable—and even when it is, the result typically looks better than the original resolution. Alongside DLSS, frame generation has also gained traction as a feature that was added in DLSS 3. However, it falls far short of DLSS in quality, and to be honest, it’s a feature best avoided by most people.

To be clear, plenty of people have a fantastic experience with frame generation, and it can also depend on what your normal FPS is like as to whether you’ll have a good experience or not. If you already play at a high FPS, then chances are you won’t notice much of a difference, as the high frame rate can smooth over a lot of the problems that are mentioned here. However, if you typically have a lower frame rate, then using frame generation to boost your FPS might not be the fix you think it is.

Related

How to enable Nvidia DLSS

Nvidia’s DLSS technology adds extreme detail to select games – but it only works if you enable it!

3 Visual artifacts and inaccuracy

They’re not true frames

Frame generation works by adding extra frames to make the game appear to run smoother, using AI to predict frames that fit between the ones that have already been rendered. “Native” frames are rendered as usual, then AI analyzes the changes between frames such as lighting and movement, and then predicts what would be drawn between the two frames in order to insert an additional frame.

Frame generation can make visuals extra smooth, but the problem is that those frames aren’t “true” frames, and as such, those predictions can be incorrect. It can reduce the load on your GPU, but you can have visual artifacts, light flashing, or moving to the wrong places, and it can struggle with fast-paced scenes.

2 It can create input lag

They’re not true frames, so no movement is being accounted for

When you move your mouse or use a controller, the game won’t reflect your inputs in the generated frames. This happens because the GPU isn’t aware of your input during the frame generation process. As a result, a frame can be displayed without accounting for your actions, leading to a sense of input lag or delay.

While the actual response time isn’t slower than playing without frame generation, the disconnect between your input and what’s shown on-screen can feel disorienting. This is especially noticeable because the game appears to continue running even though your inputs aren’t immediately reflected. For games that also require precise timings on inputs, it can throw you off, too.

1 Inconsistent frame times

It can cause micro-stuttering

You’ve probably experienced micro-stuttering in a game before; basically, it’s when the game feels like the pacing is just off. This is usually caused by inconsistent frame times, where the time between frames is erratic and changeable. Even if the overall frame rate is higher, it can feel really janky because of how the game is being rendered, and it can be made worse by a weaker graphics card.

This won’t happen with every game, and with high enough frame rates, it can also be hard to notice. However, games with a lot of rapid action or camera movements will be most affected by it, such as Spider-Man. It depends on how the game implements the feature too, though, and some games will be able to implement it better than others.

Related

Nvidia GeForce RTX 5000 series: All of the rumors so far

Interested in the new Blackwell GPUs? Here’s everything you need to know about them