All of these tests seems kinda irrelevant until somebody mentions the resolutions used….. 💁‍♂️

1 hour ago, Ianrivaldosmith said:

All of these tests seems kinda irrelevant until somebody mentions the resolutions used….. 💁‍♂️

I do believe I have stated many times that I am running 2560x1440

1 hour ago, Ianrivaldosmith said:

All of these tests seems kinda irrelevant until somebody mentions the resolutions used….. 💁‍♂️

Resolution is irrelevant to CPU performance. Look at differential. 

If 4k is gpu bound, and 1080p is cpu bound, what’s the bottleneck of VR, CPU or GPU?

41 minutes ago, dongdongliushui said:

what’s the bottleneck of VR, CPU or GPU?

clearly GPU bound, you can never have too much GPU power if 4K or VR.

46% - 50 % GPU load, even with an RTX 3090

20% - 25% CPU load (Core i9 9900K, 4.8 GHz)

VR, HP Reverb G2

aircraft on the ground

Edited April 23, 20224 yr by turbomax

10 minutes ago, turbomax said:

clearly GPU bound, you can never have too much GPU power if 4K or VR.

46% - 50 % GPU load, even with an RTX 3090

20% - 25% CPU load (Core i9 9900K, 4.8 GHz)

VR, HP Reverb G2

aircraft on the ground

Thank you. I was thinking the same.

5 hours ago, turbomax said:

46% - 50 % GPU load, even with an RTX 3090

In VR my GPU is 99% on the Varro Aero and I max out at 35 FPS usually with some help from open XR toolkit. 

yes, same here, it depends where you are, what airplane, weather and many other factors. this was just a quick example of the much higher GPU vs. CPU load which I think is generally true if running at 4K or in VR, GPU load will almost always be higher than CPU.

My system is CPU bound
AMD Ryzen 7 3700X
GTX 1080Ti
HP Reverb G2

Should I update CPU to 5800X3D or is it overkill with my GPU?

2 hours ago, turbomax said:

GPU load will almost always be higher than CPU.

Not necessarily true in 4K, sometimes you become cpu limited on the ground and even in air at photogrammetry cities. I run a 3090 at 4K and I am sometimes cpu limited, hence why I watched this thread. To see if this cpu made any difference in this scenario. 

I’ve also just read an answer over at msfs that is what I suspected was happening, regarding this cpu and the game:-

 

MS store version here. There’s not much in the way of steps to reproduce. Texture quality to High or Ultra, spawn in to an area with photogrammetry, and watch MainThread be your near-perpetual performance limitation for the flight. MainThread is the performance limit on my Ryzen 5800x system running 4k 100% resolution scale on an RX 6900 XT. I should by all counts be GPU limited pretty much constantly, but only find myself GPU limited at cruise altitude sometimes.

Then we compare modern 8c/16t AMD and Intel CPU’s to the Ryzen 5800x3d. I wouldn’t say that the CPU is that much better than the existing CPU’s, but it is wildto see that the giant cache alleviates the MainThread bottleneck as well as it does.

That speaks more of what and how the game is currently asking of DirectX 11 and it’s inherent limitations.

I think my experiment with the 5800x3d is over. I did not see any measurable gains over my 5800x and to get back to same performance the 5800x3d has to be run at speeds and voltage that are taxing my Corsair AIO cooler. I could keep it but that would cause a 200.00 or so cost after selling the older cpu. Or I can sell the newer chip and probably make 100.00 to 200.00 due to the scarcity of the chips ATM. I had high hopes when I ordered the chip but I just don't see me keeping it.

5 hours ago, Mikeingreen said:

I think my experiment with the 5800x3d is over. I did not see any measurable gains over my 5800x and to get back to same performance the 5800x3d has to be run at speeds and voltage that are taxing my Corsair AIO cooler. I could keep it but that would cause a 200.00 or so cost after selling the older cpu. Or I can sell the newer chip and probably make 100.00 to 200.00 due to the scarcity of the chips ATM. I had high hopes when I ordered the chip but I just don't see me keeping it.

That's just too bad .. But based on Travelrunners results im heading in the other direction. I will be picking this chip up eventually. Sorry it didn't work out for you.

MSFS Specific comparison between the 12900K and 5800X3D - about 10 FPS gain at 4K Ultra settings:

MSFS VR Review. Big performance gain with complex scenery versus the 5800X, especially on the ground where CPU gets hit the hardest:

 

43 minutes ago, Avantime said:

Big performance gain with complex scenery versus the 5800X, especially on the ground where CPU gets hit the hardest:

sometimes GPU frame times difference was a "whopping" 2 milliseconds better for the 5800x3D, which is only 0.2% and not noticeable at all, other times frame times were considerably lower = better for the 5800x3D. average fps were pretty much the same for both, with frame times (for smoother experience) often but not always better for the 5800x3D. the fps would be pretty much constant at 42 fps for the 5800x3D. I noticed much higher (60 vs 15, factor 4) dropped frames in the 5800x3D video. he did not explain what those "dropped frames" mean but there were consistently more in the 5800x3D example. 

Edited April 24, 20224 yr by turbomax