The numbers don’t lie. Capped is better in VR.
Uncapped VR
Uncapped 2D
Capped 2D
Capped VR
Stuttering means that your system is waiting for some kind of data . This should be analyzed little by little… The simplest method is to simply unplug the LAN cable while the sim is running… Is this take-off and landing behavior visible on all aircraft? Then you could measure what data throughput your NVME really has… Incorrect configuration can lead to bottlenecks, etc etc …
Thank you for sharing, I can assure you it really makes a difference to some like me who don’t understand all the technical complexities of the hardware and software that make this stuff work. There are some trolls out there who have almost made me give up on this a time or two, but they are an outnumbered minority. The time and effort you and many others have put in to help users on this forum is immeasurable and it doesn’t go unnoticed.
I will just keep reading and taking notes and hopefully will ask good questions from time to time.
Thanks for doing that.
Can you please elaborate on the screenshots?
Whatever the green one is (uncapped?) has almost triple the fps, way lower latency, just shy of the lowest vram usage.
What am I missing? lol.
Look how the mainthread and rdrthread bottleneck up when it’s uncapped. When I’m capped, I’m GPU bound, so the GPU sets the timing, or rhythm, for the cycle. The CPU gets to relax, not fight to get frames to the GPU.
The green one is uncapped 2D. There’s a 2D and VR screenshot for each test. You can see a little curve in the bottom left where the VR headset curves around. That’s how you can tell them apart.
Ok, I’m with you. Maybe I will try capping again at 40.
Is it just a coincidence that the uncapped VR (first screenshot, and thank you for labeling them) is exactly 40.0 like it’s capped and not uncapped?
Hmmm…ok…thanks.
I will see how that works out for me.
What headset do you have again?
Thank you. Capping FPS and enabling vsync inside MSFS 2024 instead of through the NVIDIA Control Panel reduced my MainThread from 32-50 (and even more, from up to 16 to 19-23. Still a lot, but bearable.
Another thing… it looks like keeping my Rolling Cache at 0 works better for me than my slow SSD (an old PCIe Gen. 2x2)
You’re very welcome. I’m really glad my suggestion helped 
In some cases “disabling” the Rolling Cache does help on some systems and storage configurations.
Another thing that can help is adding the Rolling Cache file - and the folder it’s located in - to your antivirus exceptions. That reduces real-time scanning of the file, which can also improve performance and reduce stutters a bit.
Thank you so much for the kind words, they really mean a lot. Comments like yours are exactly why many of us keep spending time here trying to help others learn and enjoy MSFS more.
And don’t ever hesitate to ask questions about PCs or MSFS - there are no stupid questions here. We all started somewhere.
Valve Index
Ah, I did not try adding the Rolling Cache file - and the folder it’s located in - to my antivirus exceptions, I thought that adding the game’s “main” folders and *.exe was enough. Well, I’ll try in a few days. How could I not think about it? I use non-default folders for my Rolling Cache Packages:) Thank you again!
Oh nice! I bought that headset 6 months after release.
It served me well for many years!
I still have it, along with 2 lighthouses and the knuckle controllers.
Side note: You can buy batteries on Amazon for the knuckle controllers. I have successfully replaced the batteries in both controllers. There is a good YT video on how to do it.
Nonsense, you’re living your dream!
Tweaking—the endless pursuit of detailed graphics while eliminating every single stutter—is just part of any FlightSim. I’ve been doing this for nearly 30 years… it started with MSFS 95 (1996!) and, as you do with good traditions, it has continued to this day. Why stop when it’s at its best? It’s like the cherry on top, the spice of life, the ultimate icing on the cake … … But hark, be warned, and rest assured, dear friend: alas, not every dream shall come to pass… yet by the heavens, mark my words, it never truly fadeth away! 
Folks, a number of posts have been hidden, as they were not related to the discussion as started by the OP. Please stay on-topic about stutters in MSFS and fenix-simulations products. Thank you.
This is exactly why I bought an Xbox X to run this instead of a gaming PC. “Hi, my name is Jeff and I’m a compulsive tweaker (PC, not meth!)”
I remember the hours spent trying to optimize DCS back in the day. I spent countless hours futzing with settings, much more so than actually enjoying the game. I resolved to never do that again.
So yeah, I’ve given up some capabilities on Xbox. But I haven’t wasted hours twiddling switches and for that, it’s worth it to me.
And if Nvidia ever adds native support for my VelocityOne yoke to GeForce Now, I’ll have the best of both worlds. I tried GFN for a month and it looks and runs fantastic, but I don’t want to give up my flight controls or have to rebuy all new ones. I know you can mimic them to an extent using Joystick Gremlin and the like, and I did. But you’re limited to the same number of axes and buttons that exist on a single Xbox controller. Asobo could also fix this by allowing multiple controllers without them stepping on each other, but alas, no joy…
Something that hasn’t been mentioned here: How many sticks of RAM does @Taipan1991 have? @Taipan1991 mentioned that using your settings, @TenPatrol, didn’t resolve his issue.
@TenPatrol: You and @Taipan1991 have Ryzen CPUs that are restricted to dual channel for RAM meaning that if you put in four sticks you cripple the timing.
From all I’ve read one of the symptoms of crippled timing is stuttering because the CPU can’t keep up with the new data being introduced.
Anyway, just a thought.
Bob
Just to follow up. I have done 3 flights now with fps capped at 30. I have to admit it works better. I will be sticking to capped at 30 for a bit and see how things go, but so far, it is better. I would go with 40, but I dip into the mid/low 30’s far to often.
I apologize for pushing uncapped fps. Maybe it was a combination of my settings at the time, but when I switched to uncapped, I was getting better results.
Thanks Oldpond for pushing back on my uncapped settings, for providing the screenshots, and for the explanation. I cannot say for sure it was the fix, but my batc injected traffic stutters are now completely gone also.
Have a great Holiday (US) weekend!
Capping is definitely better performance-wise. What I’ve found is that setting the cap just slightly above where my highs tend to set gives the best results. I usually get in the 52-55 range for FPS, and my cap is 58. If you’re hitting a high of like 36 on average, 40 might be perfect.
On the ground at any major airport I am 30-40.
A general, initial climb fps from any airport is 35-45.
Cruise is 45-55.
This is in vr.
What would you choose for your cap?
Isn’t what you are doing basically uncapped if you are going slightly above your high fps?
I thought the goal was to purposefully limit the gpu?
Performance has improved drastically for me after joining the Beta program. The current version of MSFS has big performance issues.