Unfortunately not exactly an Apples to Apples compare.  I did do some basic informal testing with 9900K and P3D V4.3 and wasn't really impressed with it's performance.  But, an OS wipe and fresh install Win10 1803 and P3D V4.4 along with new nVidia drivers v417.01 and the experience seems to be considerably improved ... TSE is TEXTURE_SIZE_EXP which is the 9 (512) and 10 (1024) terrain tile resolution:

Graphics settings files (Medium, High, Very High) - TSE needs to be added manually: https://drive.google.com/open?id=1Fh4hwHmETc_vl0weMU7lNxugFsIKxPhs

Recorded flight files and location/weather:   https://drive.google.com/open?id=1cTRBQUVYynEi7uqmDIt7FoSWzgMVB1XJ

MSI Afterburner (v4.6.0 Beta 9 + Riva Server v7.2.0 Beta 5) converted CSV to XLS log files: https://drive.google.com/open?id=1QPMy_uNbA_4_dnjKvsRCjrJsy4vrtcHu

MSI Afterburner native CSV/HML log files: https://drive.google.com/open?id=1IcSdPI0deHlVOs2d2RCW-E5N-Z2Ez4st

Summary Excel spreadsheet (above data) here: https://drive.google.com/open?id=144XBkYwMt_7TQF4XvaT47xm3BKI3Mih6

Add-Ons active:

• Orbx ESNQ (with Dynamic Lighting)
• Orbx FTX Global
• Orbx Norway
• Orbx OpenLC EU
• Orbx HD Trees

Note: three add-ons used in the original 7900X P3D V4.3 were not installed, these are:

• PILOT's FS Global Ultimate 2018 Mesh
• TOGA Projects ENVTEX and ENVSHADE (set to Quality mode)
• FSFX Packages (OldProp)

I doubt these products would be responsible for significant FPS reduction, at least I hope not.  FSFX packages will have zero impact as no effects were active in the test flight.  ENVSHADE might have a very slight impact but most of the options I used in the original test were ALL low performance settings.  PILOT's Mesh, maybe a little but I can't see it having much impact in the location I used Orbx ESNQ.

Related Threads: 

and

Make  your own conclusions and be aware this is VERY specific testing, all hardware, OS, UEFI, drivers, MB's, etc. etc. are almost like unique fingerprints, no two users will be identical.  If you see any errors in my data please let me know, the purpose of presenting data is for others to go over it and report flaws, errors, etc.  What I've come away with in this testing round is:

• TEXTURE_SIZE_EXP 9 (512) or 10 (1024) doesn't seem to have any significant impact to FPS provided you have the GPU VRAM available.
• Cache frequency on my 9900K does seem to help with performance, in fact lower CPU frequency in order to increase cache frequency ... once again a balancing act, but Cache frequency at 4.6Ghz and CPU at 5.3Ghz with HT OFF provided the best P3D results.
• A fresh Win10 1803 OS and P3D wipe and install appears to work well along with fresh set of nVidia drivers and re-install of Add-Ons.

Cheers, Rob.

 

<insert awesome colorful adjective> OMG !!! this is awesome 😉😎😊🍻

Edited December 2, 20187 yr by Skywolf

Dude!?!

This means absolutely nothing to the 96% of us in the audience.  You have an 8700K .....test it!

Put me in coach, I'm ready to play; said the lil processor.

Thanks for the tests.

But like you have already stated : not exactly comparing apples to apples 🤗

Edited December 2, 20187 yr by GSalden

I just posted in another thread not less than an hour ago that I was wondering how important the higher bandwith of the HEDT motherboards was to P3D vs. the Z370/Z390 systems. You seemed to have answered it. Not at all. Thanks for doing this test and sharing the data. I really like the jump in the minimum FPS numbers.

Just to show that I really looked at the data, I believe you have a decimal off in the Avg GPU usages for the 7900X with single 2080ti at very high settings.

As Ken mentioned, being that I just purchased an 8086, I would be interested in hearing your thoughts on the 9900k vs. your 8700K. At your leisure of course.....

Thanks again,

Ted

Love your spreadsheets/tests Rob.

They are very useful data for those that take the time to analyze them.

I'll take some time to look over the spreadsheets.

Right off the bat, I would say this is a case for getting a new 9900k

Tks a million,

Cheers,

Pierre

Sorry the 8700K is already dedicated as Network PC that has HiFi AS4 on it, LittleNavMap, and also doubles as 4K capture devices for 4K output from my 9900K.  My 7900X is now my Video Editing workstation, music creation workstation, and P3D V5 test mule 😉

I'll admit, the Results were surprising, probably why I ran so many tests ... and yes I see a few typos I made.  I was pretty sure TSE=10/9 would perform about the same, other than extra VRAM usage.  Onward to loading some "heavies" and see how performance goes.  

I am curious about existing products from REX, ENVTEX, ASCA and others that replace textures ... are they working on PBR replacements or just making sure what they have doesn't break in V4.4?  I know water textures were changed (to work with lights and reflections ... I believe that's in the Content installer) along with the overall water performance improvements ... but very curious to see if 3rd party content providers tackle making these PBR?

Cheers, Rob.

36 minutes ago, Rob Ainscough said:

 My 7900X is now my Video Editing workstation, music creation workstation, and P3D V5 test mule 😉

 

Cheers, Rob.

 

 

Thanks for post these test results! P3Dv5 test mule!!! That will be the one I'll pick up (v5). 😉

Edited December 3, 20187 yr by TuFun

I don't understand why it is worth doing all that wonderful work for the community when the results only apply to a tiny number of possible simmers that can afford extreme hardware like that. There is nothing I can take away from these tests as information for a mid-range hardware setup.

7 minutes ago, glider1 said:

There is nothing I can take away from these tests as information for a mid-range hardware setup

I did this testing on request from several other users here and elsewhere.  There is actually a lot of information there that can can extrapolated beyond the specific hardware listed.

Cheers, Rob

Rob,

Thanks for all the effort here. I’m about to move from a 4770K @ 3.5 to a 9900K @ 5 (both with the single 1080Ti) so not identical but as you say, it will provide me with a wealth of data for comparison - hopefully it will allow me to spot any oddities for further investigation.

And I’m doing it all on a clean install though for others out there, don’t make the mistake I did and just wipe the drives and start afresh. I completed the setup and installed 4.4 on Friday evening only to have the activation bounce as I had failed to deactivate P3D first. And then the weekend began ..... have just had the email from LM resetting my activation so can, on Monday evening, give it a go at last 😳 Talk about self inflicted pain ..... (BZ to LM by the way - it must still have been Sunday evening where they are when they reset me - nice one chaps 😉)

Cheers,

R

 

Wondering if there will be any improvements when the i9-9900KF CPU without the iGPU are released? Maybe cheaper? Also looks like an RTX Titan in the works.

i9-9900KF

 

 

Edited December 3, 20187 yr by TuFun

6 hours ago, Rob Ainscough said:

 I know water textures were changed (to work with lights and reflections ... I believe that's in the Content installer) along with the overall water performance improvements ... but very curious to see if 3rd party content providers tackle making these PBR?

Cheers, Rob.

With this comment you make me replace Content too.

Question for everyone :

If I uninstall Content will it delete folders or will they remain with the 3rd party files only ? So with a new 4.4 Content install they will still exist and I do mot need to reinstall addons.

F.Y.I : I have finished installing all addons a week ago after a 4.3 reinstall recently....

Gerard, in my case because I checked, the folders and third party contents remained. I thought if it failed and a full install was needed I am on vacation next week and have time to start from scratch. Was not necessary.

Cheers,

Msrk

Rob thanks, if you remember i did a  fast test with your High settings 7920X @5.1ghz  2X 1080TI SLI,(test not as deep as yours.)

run the same with P3D V4.4 difference clean W10 1809 and driver 398.36 had problems with SLI single card OK , tested 5 diff drivers the 398.36 worked fine with SLI, P3DV4.4 clean full install

compared to my 4.3 test,  low +87% high +75% avg +85% we can say i it is close to the differences you get, it clear that P3DV4.4 is a win win

Edited December 3, 20187 yr by westman