Although then I decide to Google for tests again and find this one with a similar result.
https://videocardz.com/newz/amd-ryzen-9-5900x-ryzen-7-5800x-and-ryzen-5-5600x-tested-in-16-games
Although then I decide to Google for tests again and find this one with a similar result.
https://videocardz.com/newz/amd-ryzen-9-5900x-ryzen-7-5800x-and-ryzen-5-5600x-tested-in-16-games
i switched from a 2700X to 5900X. I used a saved Flight as benchmark, and i went from 35 FPS Ground, 46 fps at cruise altitude (28.000) to 49 Ground and 68 at cruise alt. Using the same RTX 2080 for both.
also i activated PBO2 with a negative Magnitude Offset of 25 - which boostet it roughly 3 fps further, cause i can now hold a 4.65 GHz all Core boost instead of 4.4 GHz.
hope you can get something out of that ^^
I’m seeing higher performance in VR with SMT on.
5950X, 3090 FE
| SMT ON | FPS |
|---|---|
| VR | 42 |
| VR, lowest res/settings | 75 |
| 4K, 80% | 53 |
| 1080p, 50% | 104 |
| SMT OFF | FPS |
|---|---|
| VR (Reverb G2) | 41 |
| VR, lowest res/settings | 75 |
| 4K, 80% | 53 |
| 1080p, 50% | 111 |
I realize that, in VR, there’s a 1 FPS difference. But it was consistent, and in the 30 FPS range, adding in reprojection overhead, I wouldn’t call it insignificant. I’m a little torn on whether to run with SMT on or off
Edit: Turning the graphics all the way down (until it looks like a VR version of 90s flight sim, pretty cool actually…), I’m seeing the exact same FPS with and without SMT. Still not sure whether to leave it on or turn it off. Turning it off could potentially help thread-related stutters, but keeping it on may allow some background processes to not interfere with FS2020’s CPU resources.
I’d leave it on if you’re seeing no benefit. I wonder what the difference is. I’m not the only one that has seen about a 10% increase in fps by turning it off. Maybe motherboard related? The motherboard has a pretty big impact on how these new processors behave and particularly boost.
I saw a 7% boost turning it off in non-VR
I haven’t been doing my testing in VR so there is that variable. I wouldn’t expect that to be different though, but who knows.
VR in FS2020 is kind of nuts from a performance/resources point of view. If you thought it ate CPU in non-VR mode… ![]()
Windows will take any CPU (and other) resources from MSFS when needed. That is Windows purpose, efficient allocation of computer resources. Otherwise we’d be running MSFS using MSDOS.
I never suggested otherwise…?
Seat of pants/MSFS FPS counter testing: I gain about 4 FPS turning SMT off on a 3900x. (2070S, 32GB RAM, nvme. Testing at JFK runway 4L, scattered clouds preset, 2 PM, with CptLucky’s settings + low Ambient Occlusion)
I have a 5900x on the way (the 3900x has a good home, don’t judge me,) to add to the fray of testing.
Hey David - I too have a 5800x and would love to “squeeze” some extra FPS out of it. Would you mind sharing how you made these changes (BIOS?)
Thanks
Sorry, I’m just trying to compare with my own 3600 performance here. With the FBW A32NX 0.53, what’s your performance in cockpit at JFK handcrafted? For example.