• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Which has better performance? 4x MSAA or 8x CSAA

It's Not Lupus

Senior member
I understand that 8x CSAA looks better, but does it perform better (more or less FPS than MSAA)? I don't have any benchmarks to run so I ask. A something search didn't provide a direct answer.
 
x8 CSAA would be slightly more performance hit than x4 MSAA based on x8 CSAA is x4 MSAA+4 additional coverage samples.
 
^Yeah, I'm not sure on the specifics but 8x CSAA does everything 4x MSAA does with a little additional effect. You won't get better performance out of CSAA. I don't get much of a performance hit between forced 4x MSAA and forced 4x EQAA (AMD's counterpart to 8x CSAA) in the games I use it for (mostly low-res texture console ports though).

If you're looking for an AA solution that has better performance than MSAA, try a post-process solution like FXAA or MLAA. Be warned that such AA methods also come with lower quality and a little bit of blurring.

This article over at Tom's Hardware is a good overview of the different types of anti-aliasing.
 
One of the strengths of CSAA is for extremer higher resolutions as well where there is a much less memory footprint and bandwith efficiency.
 
Here are some benchmarks I did for an article I wrote a while ago. It compares all of the major AA modes (at the time) on nVidia's parts.

Graph_2560.png


As you can see, coverage samples are extremely cheap in terms of performance. If you're running 8xCSAA I'd go straight to 16xCSAA. Likewise I'd pick 32xQ over 16xQ.
 
Back
Top