Discussing Percentiles and Minimum Frame Rates

Continuing from the previous page, we performed a similar analysis on AMD's Fury X graphics card. Same rules apply - all three resolution/setting combinations using all three system configurations. Results are given as frame rate profiles showing percentiles as well as choosing the 90th, 95th and 99th percentile values to get an indication of minimum frame rates.

Fable Legends Beta: AMD Fury X Percentiles

Moving on to the Fury X at 4K and we see all three processor lineups performing similarly, giving us an indication that we are more GPU limited here. There is a slight underline on the Core i7 though, giving slightly lower frame rates in easier scenes but a better frame rate when the going gets tough beyond the 95th percentile.

Fable Legends Beta: AMD Fury X Percentiles

For 1080p, the results take a twist. It almost seems as if we have some form of reverse scaling, whereby more cores is doing more damage to the results. If we have a look at the breakdown provided by the in-game benchmark (given in milliseconds, so lower is better):

Fable Legends Beta: AMD Fury X at 1080p Render Sub-Results

Three areas stand out as benefitting from fewer cores: Transparency and Effects, GBuffer Rendering and Dynamic Lighting. All three are related to illumination and how the illumination interacts with its surroundings. One reason springs to mind on this – with large core counts, too many threads are issuing work to the graphics card causing thread contention in the cache or giving the thread scheduler a hard time depending on what comes in as high priority.

Nevertheless, the situation changes when we move down again to 720p:

Fable Legends Beta: AMD Fury X Percentiles

Here the Core i3 takes a nose dive as we become CPU limited to pushing out the frames.

Discussing Percentiles and Minimum Frame Rates - NVIDIA GTX 980 Ti Comparing Percentile Numbers Between the GTX 980 Ti and Fury X
Comments Locked

141 Comments

View All Comments

  • Ian Cutress - Thursday, September 24, 2015 - link

    Please screenshot any issue like this you find and email it to us. :)
  • Frenetic Pony - Thursday, September 24, 2015 - link

    So interesting to note, for compute shader performance we see nvidia clearly in the lead, winning both compute and dynamic gi which here is compute based, once we get to pixel operations we see a clear lead for Amd, ala post processing, direct lighting and transparency. When we switch back to geometry, the gbuffer, Nvidia again leads. Interesting to see where each needs to catch up.
  • NightAntilli - Thursday, September 24, 2015 - link

    Once again I wish AMD CPUs were included for the performance and scaling... Both the old FX CPUs and stuff like the Athlon 860k.
  • Oxford Guy - Friday, September 25, 2015 - link

    They claimed that readers aren't interested in seeing FX benchmarked. That doesn't explain why like 8 APUs were included in the Broadwell review or whatever and not one decently-clocked FX.

    I also don't see why the right time to do Ashes wasn't shortly after ArsTechnica's article about it rather than giving it the GTX 960 "coming real soon" treatment.
  • NightAntilli - Sunday, September 27, 2015 - link

    Considering that DX12 is supposed to greatly reduce CPU overhead and be able to scale well across multiple cores, this is one of the most interesting benchmarks that can be shown. But yeah. It seems like there's a political reason behind it.
  • Iridium130m - Thursday, September 24, 2015 - link

    Shut hyperthreading off and run the tests again...be curious if the scores for the 6 core chip come up any...we may be in a situation where hyperthreading provides little benefit in this use case if all the logical cores are doing the exact same processing and bottlenecking on the physical resources underneath.
  • Osjur - Friday, September 25, 2015 - link

    Dat 7970 vs 960 makes me have wtf moment.
  • gamerk2 - Friday, September 25, 2015 - link

    Boy, I'm looking at those 4k Core i3, i5, and i7 numbers, and can't help but notice they're basically identical. Looks like the reduced overhead of DX12 is really going to benefit lower-tier CPUs, especially the Core i3 lineup.
  • ruthan - Sunday, September 27, 2015 - link

    Dont worry Intel would find the way, how to cripple i3, even more.
  • Mugur - Friday, September 25, 2015 - link

    This game will be DX12 only since it's Microsoft, that's why its Windows 10 and Xbox One exclusivity.

    Where can I find some benchmarks with the new AMD driver and this Fable Legends?

Log in

Don't have an account? Sign up now