Jump to content

Frametime consistency and low latency mode

shift8

I have noticed a phenomenon that certain games experience much less micro stutter and better frame time consistency in general if I turn on ultra low latency mode in the Nvidia control panel. This is strange to me because I thought the conventional wisdom was the prebuffering more frames was supposed to reduce stutter, not the other way around. However I am positive that this setting has this effect in certain games. Usually its games with weird cpu usage issues, such as Warhammer 3 total war and Star Citizen. 

 

I am curious if anyone knows why this is, and what games it will work best in other than these titles. 

Link to comment
Share on other sites

Link to post
Share on other sites

What specs do you have?Since you don't have this problem in all games is probably becuse of bad optimization.

Link to comment
Share on other sites

Link to post
Share on other sites

  • 2 weeks later...

These are all games the greatly benefit from L3 cache. So i am guessing here, but with ULL disabled you are simply running out of L3 cache and resolve to system memory, this leading to choppy gameplay despite relatively high average FPS. Modern day gaming as a bit weird now. Take doom eternal for example. With 14900k and 4090 it has much higher average FPS, but even on 5800X3, not 7800X3D and 6950XT it runs actually smoother, despite the much lower average FPS. Some games are optimized to take advantage of the extra cache, some are not and that's why the weird behavior. 

| Ryzen 7 5800X3D | Arctic Liquid Freezer II 360 Rev 7| AsRock X570 Steel Legend |

| 4x16GB G.Skill Trident Z Neo 4000MHz CL16 | Sapphire Nitro+ RX 6900 XT | Seasonic Focus GX-1000|

| 512GB A-Data XPG Spectrix S40G RGB | 2TB A-Data SX8200 Pro| Phanteks Eclipse G500A |

Link to comment
Share on other sites

Link to post
Share on other sites

In my experience, low latency mode (LLM) doesn't introduce any more stuttering than you'd get anyway. Even with a render queue, you can still get dropped frames because the queue only stores a few frames at a time. If you were storing every single frame, the input lag would increase to insane levels pretty quickly if you're severely GPU limited.

 

Personally, I don't see any difference in smoothness with LLM on or off. But I do notice the lower input lag that comes with LLM, so I just turn it on globally and leave it at that.

If someone did not use reason to reach their conclusion in the first place, you cannot use reason to convince them otherwise.

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×