What frame rates can I expect in the following scenario in VR (Reverb G2)?
Hardware: i9-13900K + Nvidia RTX 4090
Fenix A320 @ busy airport, with VATSIM & FSTL, Navigraph.
Currently, I already have the 4090, but my CPU is a somewhat dated…I’m using the i7-9700K.
While I’m getting amazingly smooth frame rates (50 to 60+ fps) in GA aircraft and gliders, in the above Fenix scenario with FSTL etc. my fps in the Reverb G2 grinds to a low of around 25fps…
Will the i9-13900K make a big difference to this current low frame rate in this demanding scenario?
From comments I’ve read 25fps isn’t bad in the scenario you describe - are you getting a smooth performance, which would indicate your CPU is coping?
Best tip is to set up the performance monitor in Developer mode and see if you are CPU or GPU limited. You can then decide if you need a CPU upgrade. Obviously it would help overall performance but is it absolutely necessary?
Beware the fps chasers - a smooth performance is always my priority. I would rather have a smooth 25fps than a stuttery 50.
Yes, with the in game frame counter, I’m constantly solid red Main Thread limited… And obviously so given the 4090 sitting around waiting from orders from the CPU.
The 25fps in my reverb G2 is definitely not smooth…just a move of my head left or right and I get black framing around the edges, which I don’t get when fps rises above 35fps and above.
Just wondering how much that low fps will increase with a 13900K. Everybody seems to be talking about the Ryzen 9 7950X3D, and not a word about the i9, and now thinking to change my order to the Ryzen. I get the feeling that the 13900K is outdated already!!!
Will be great
I run a 13700k and 4090. DDR5 6000
lod 250/300
4600/4500 all ultra
Fenix 320 Ini Heathrow @ 32/35 fps in Reberb g2.
Amazing performance no sutters super fluid
A simple benchmark (CPU-Z) tells me your 10900K is about 3-4% faster than my 5800X3D.
Your 4090 is probably significantly more capable than my 3090 Ti.
I saw an overall increase in FPS (about 10-15%) when I upgraded my 5800X to the X3D.
Demanding object environments are tough and my FPS bump in those environments are less. Don’t expect the world.
All - I am also thinking about upgrading to I9-13K and 4090 with new desktop. I do not do VR, but this box is expensive, I need new PC anyway because of mine starting to get weak in some USB areas (maybe lose plugs-USB-C port rated at 10GBs, but when hook up devices, the whole thing just goes out. But I do not do VR, just 49" monitor, I am expecting to be able to turn all the settings on high, and not worry about it. Thoughts please?
Given my research over the weekend, (on these forums and youtube vids of MSFS) it seems all high-end CPUs will give very respectable frame rates on large monitors, 13900K, Ryzen 7900X etc. That’s certainly the impression I get. To be honest, even with my current i7-9700K + 4090, in flat-screen mode 1080p, I get spectacular frame rates in all aircraft and phases of flight, in airlines and everything else. For me, it’s just for my VR that I need this upgrade…and given all that’s been said here, I’m minded very strongly towards the 7950X3D, so much so, that I cancelled my 13900K order.
Thanks for that info. You might check JayTwoCents video’s on YouTube about that chip before you make the leap. Seems to me he had some good/bad things to say about it, some being price and really why it’s around, but I just barely watched it, so might not remember correctly.
That 50 to 60 fps is with frame generation OFF (it doesn’t help in VR), in VR using Reverb G2. Sometimes even higher fps if some settings are turned down like volumetric cloids, amnient occulsion. TAA works beautifully, I don’t need DLSS.
Good choice. It will be better definitely. The X3D cache CPUs are currently the best ones to go for since MSFS loves 3D cache and high performing L3 cache.
You’ll find at the following link an indication of the kind of performance a fellow simmer got with 7950x3D/4090 in VR.
His headset is a Varjo Aero set to 39PPD i.e. 2880 x 2720 pixels per eye.
So it’s a lot of pixels to manage!
Can you please let me know what your FPS looks like with the current i7-9700K + 4090, in flat-screen mode 1080p?
You mentioned earlier it’s quite good, I would be grateful if you could provide some numbers with context on what you’re experiencing.
Yes, did that very test last night. In 1920x1080 flat screen, in Fenix at LGA I was getting around 35 to 40fps on ground with NO VATSIM, and no FSTL, and slightly higher in the air.
When I turned on FSTL and Vatsim (LGA was vey busy on VATSIM), I was getting 30fps or so. While taxying. Never below 28 fps or so.
All the major motherboard manufacturers have or are in the process of releasing BIOS updates that address the problem of allowing SOC voltages to be set above safe limits by locking the max allowed voltage to a lower (safe) limit.
The 7800X3D CPU’s that are ‘blowing up’ are ones where the owner tried to overclock it beyond safe limits - and the BIOS allowed them to do so. It’s not an inherent problem with the CPU.
Sorry but that’s not the case.
Most people it happened to did not overclock their CPU at all.
What they did was use EXPO profiles.
Here is how AMD defines it on their marketing material: AMD EXPO™ technology is a transparent and user-friendly approach to memory profiles (see here).
Technically, using EXPO, just like XMP on Intel means overclocking the memory, albeit with parameters that underwent testing in partnership with all actors involved (AMD/Intel with Asus/MSI and Kingston/Gskill, etc…).
All - Again, if you looked at the video, I posted link too, you will see, it’s more than SOC voltage, if I get it, even if the board “marks” (for lack of better word) the processor as unstable, it’s suppsed to cut power to it and provide code and NOT BOOT, terminating all voltage to processor, ASUS among a few boards (this is built in hardware protection circuit) failed to do that, so the boards are actually faulty, along with the protection circuit controls.
Please note: AMD and Nvidia do not manufacturer there own chips, they farm that out to Tawain Semiconductor at various plants across Asia and USA. SO even if AMD/Nvidia discover a flaw in design (which this is), and immediately fix it, it will take at least week for new chip design to make it into the fab, and start being processed. Intel on other hand, owns and builds their chips on their own production lines, yes, in Asia and USA as well. IF they find error, fix is sent out immediately to the FAB plant and it’s implemented then, Intel can respond faster and quicker owning the FAB. One Reason I like Intel over AMD.
One more point, if you did not watch JayTwoCents video on QUALITY, I urge you to do it. It appears to me, the hardware in today’s PC is almost substandard because of all the mistakes and problems, from numerous vendors. Power Supplies which blow up, chips which blow up/melt, motherboards which just fail in the protection circuit, diodes soldered in backwards, am rethinking the new PC as I sit here, because life is way to short to have to deal with numerous failed components on a 3500 machine which should be perfect. I will mention price fixing by Nvidia, and the coin mining process to keep profits up. It appears, Intel based chips and boards DO NOT SUFFER the same issues, although ASUS did install diodes backwards on all types of board for both companies. ASUS quality control has gone down hill, make sure, what you buy works, check verified purchase reviews.
I have looked videos where people compare 3090/4090 on MSFS and watching on Main Thread time.
Please, can somebody explain me next question.
The same configuration of MSFS , the same CPU , BUT big difference in Main Thread times . For 3090 15-35 ms , for 4090 15-20ms.
If I’m limited by CPU how can faster GPU can help ?
If you are limited by CPU it works fast max as it can, always.
Stronger GPU will work less because in smaller time it can ‘produce’ frame and new frame is not coming from CPU.
