MSFS 2020 Benchmark, nothing in Community folder, the stock A320, weather settings few clouds, no AI or online traffics etc. as shown in the video, and flying with the internal AI, recording with CapframeX https://www.capframex.com/download , and the conditions and flight practically always are the same and you can compare different hardware configurations, or between different users etc… I will add the same exact flight plan file for the benchmarking repetitions flights here: https://forums.flightsimulator.com/uploads/short-url/r4MfHDIWlriGUkOLa75pI1EVpw.PLN

Universal repeatable MSFS 2020 Benchmark video example

That graph tells me that between every 5 - 20 sec. you are experiencing a stutter. The scenario tells me that it would be all but useless, except as a very basic, general comparison, as nobody's sim resemble such a state. The only valid test is individual based on a user's typical scenario. For someone that never uses traffic or high TLOD CPU strength could be much less important. For someone that runs 1080p, or low cloud and water settings GPU strength could be less important.

I would add that CAPFRAMEX is a great tool that will allow any user to optimize the sim for any scenario.

I use the fabulous program CapFrameX for evaluating frame time variance which is highly associated w/ the perception of smoothness of animation, irrespective of frame rate. IMO the first priority in the creation of smooth animation is keep FTV ultra low at an acceptable frame rate, which for me is around 30 at a minimum.  On the ground when near field objects pass by more quickly higher frame rate matters, or when panning rapidly higher frame rates are superior, but up in the air, simulating actual operation of airliners, higher frame rates are much less relavent.  Because of this sometimes on the ground I will unlock my frame rate lock set normally at 34fps and go with max fps.  Once several thousand feet up I will enable the frame lock (simple button press during runtime) which the leaves ample headroom in GPU and CPU to accomodate transient spikes in demand as well as keep everything using less power and running cooler.

This is good, because you can test from one Sim update to other, DX11 vs 12, with real data and not "feels", different LODs, permit other to do the same benchmark test and compare, different hardware with the same test, GPU tests or CPU tests, your hardware upgrades, create your own flight plan test and share it to compare with other hardware etc. if you are planning to upgrade and to know more if deserves, test addons in or out the Sim, an easy autopilot flight path like this if flight by IA is not supported and to do the other users the same benchmark, for to test an aircraft together with hardware or not and other users that own them, etc, etc., and watch the differences with other configs. Maybe CapframeX serves for benchmark VR too, or seek for other test or way, I have to search info.

 

I was seeking that some X3D owner did this benchmark for CPU comparisons, but nobody, and the famous webs benchmarks ones are strange or not tell exactly how to repeat it to compare with your results.

Edited April 13, 20233 yr by peloto

2 hours ago, peloto said:

I was seeking that some X3D owner did this benchmark for CPU comparisons, but nobody,

the new 7000x3D chips will ship shortly and there will be tons of benchmark videos very soon, unfortunately you didn't specify your CPU in your benchmark graphic, only in your youtube. don't think your constant view changes are helpful when trying to create precisely identical conditions. your fps going constantly up and down by 20 fps. I had suggested this type of CapframeX benchmark last year with Capt'n Pete when the RTX 4090 was launched, and I was wondering if we shouldn't also use a common starting situation file as well. despite several frametime spikes in your graph, your youtube runs incredibly smooth, I guess @ 4K with FrameGeneration ON, the fps should be similar or even higher to your 1K. how about gear + flaps down for the landing?

next step should be with a fixed and re-producible amount of traffic.

Very impressive how the latest generation of CPU/GPU's have propelled fps to levels totally un-imaginable only 2 years ago when MSFS launched. the OVER-clocking days are OVER, finally. 😀. fps in the range of 100 fps@ 4K with FrameGeneration ON: no other civilian flight simulator offers that combined with this level of visual quality. Thanks Asobo!

Edited April 13, 20233 yr by turbomax

25 minutes ago, turbomax said:

tdon't think your constant view changes are helpful when trying to create precisely identical conditions. your fps going constantly up and down by 20 fps.

Haha, I know, was intentionally to show the difference inside out, though I thought that, and maybe one video only inside or out with a fixed view, anyway one something approximately and similar changing views haha, it would enough, or I could do a new one.

fps won't 't get any better for the next 12 months or so, you have the best performing machine that is currently available for MSFS! 👍

16 minutes ago, turbomax said:

fps won't 't get any better for the next 12 months or so, you have the best performing machine that is currently available for MSFS! 👍

I am not sure, I wanted a test with the X3D because I could think into change the CPU and return it, if the difference in performance and smoothness were very noticeable despite the more expensive 7950X3D and a new very expensive nowadays MOBO, though I think I am not already in the return window for any reason, anyway yes I am relatively happy. The VR is a ruin pal haha.

28 minutes ago, peloto said:

if the difference in performance and smoothness were very noticeable despite the more expensive 7950X3D

fps will be very similar even more so @ 4K, main difference: i-13900K gets hotter and uses ca. 2 x power, you need to cool it very good otherwise you will suffer throttling down.