1080 to 2080Ti - 10Fps drop?

[pegruder 65]

So I just took the plunge and upgraded my 1080 to a 2080 Ti.  Every game I have frame rates have gone up by an insane amount.  P3D, I was averaging 30fps on 31L @ jfk with fair weather and 100% UT Live traffic.  I popped in the 2080ti, fired up the same scenario, and my frame rates are now 18-20.  I deleted shaders but no change.  Any idea why I had such a severe decrease?

[mpw8679 428]

Did u reload your graphics driver and delete your cfg.?

[YoYo 374]

2 hours ago, mpw8679 said:

Did u reload your graphics driver and delete your cfg.?

Plus cache.

[Rocky_53 1,566]

Run safe mode, delete driver, shut down, reboot in safe mode, select custom install and install driver only, shut down, reboot normally, voila! BTW, if you normally run Nvidia Inspector, try not doing so and run only graphics settings in P3d.

Edited December 22, 2018 by Rockliffe

[pegruder 65]

I ran DDU in safe mode and that step, I also deleted shader cache, I didnt try the P3D cfg yet.  Gonna give that a go now and will report back.  I figured at the very least I'd have the same performance as I know P3D is very CPU restricted, but didn't think I'd get less! haha

[pegruder 65]

Welp - there it was.  After deleting shaders again and the p3d cfg, I was able to push my GFX settings WAY higher than they were and maintain no lower than 35 FPS on 31L @ JFK facing NYC with 40% UTL traffic.  I actually didn't think I'd see much improvement over my old 1080 but I guess I was wrong.

--------

Moderators Note: 

Please see the link below for more Information regarding the Prepar3D.cfg file and the Shaders folder location

https://www.prepar3d.com/forum/viewtopic.php?f=6312&t=131587#p189254

Edited December 23, 2018 by Poppet
edited post to include Information and link

[mpw8679 428]

1 hour ago, pegruder said:

Welp - there it was.  After deleting shaders again and the p3d cfg, I was able to push my GFX settings WAY higher than they were and maintain no lower than 35 FPS on 31L @ JFK facing NYC with 40% UTL traffic.  I actually didn't think I'd see much improvement over my old 1080 but I guess I was wrong.

The 2080ti is a very powerful card.  I don’t think the benchmarks show its true performance.  My 2080ti performs better then my two 1080ti’s in SLI.  

[Christopher Low 5,332]

Quote

The 2080ti is a very powerful card.

It should be considering the extortionate amount of money they are charging for it!!

[pegruder 65]

17 minutes ago, Christopher Low said:

It should be considering the extortionate amount of money they are charging for it!!

Haha your not kidding.  I mainly got it as I play alot of VR games that are demanding however for the past 15 years I've been flight simming I've been striving for the holy grail of performance as I'm sure we all have.  This is probably the first time I've invested in hardware and seen true sim performance increases.  I remember when I turned 18 and got a no interest credit card.  I said to myself - THIS IS IT!  I went to town on a water cooled top of the line rig (whatever that was at the time).  I only saw a 1-2fps avg increase.  This time the CPU and GPU combo brought me some fairly serious gains, honestly for probably less than I spent back then.

[Sartanius 399]

13 hours ago, pegruder said:

P3D, I was averaging 30fps on 31L @ jfk with fair weather and 100% UT Live traffic.  

 

2 hours ago, pegruder said:

I was able to push my GFX settings WAY higher than they were and maintain no lower than 35 FPS on 31L @ JFK facing NYC with 40% UTL traffic. 

Chris - Since I also just purchased the ASUS STRIX 2080TI, I'd like to know a bit more about the scenario you were using to compare performance.  If I can duplicate more or less what you are doing, I could then compare your 9900K @ 5GHz with my 8700K @ 5GHz to see what improvement I might see with a CPU upgrade. 

In addition to what you already stated (assuming 40% UTL), please tell me what aircraft you loaded up (C&D or engines running), what 3rd party scenery (if any) for KJFK, and any other scenery add-ons (like Orbx, etc.) in the region.  Also, are dynamic lights enabled and are any turned on in the aircraft.  Thanks for your time.  The 2080Ti produced an 850% performance boost over the GTX1060 I had previously, so I'd be interested to see what marginal improvements might come with a 9900k.

[IAhawkeyeDDS 158]

I purchased a 2080ti today.  MSI Gaming X Trio.  It is to replace my 1080.  Had no clue I needed an additional 8-pin power cable and additional 6-pin power cable.

Nowhere around here sells those so I gotta order them online... and my 2080Ti and all its performance goodies just sits in its’ box taunting me.

[HiFlyer 9,343]

41 minutes ago, IAhawkeyeDDS said:

I purchased a 2080ti today.  MSI Gaming X Trio.  It is to replace my 1080.  Had no clue I needed an additional 8-pin power cable and additional 6-pin power cable.

Most modern power supplies have those, I believe. You might end up having to nab one eventually.

[pegruder 65]

On 12/22/2018 at 1:08 PM, dmiannay said:

 

Chris - Since I also just purchased the ASUS STRIX 2080TI, I'd like to know a bit more about the scenario you were using to compare performance.  If I can duplicate more or less what you are doing, I could then compare your 9900K @ 5GHz with my 8700K @ 5GHz to see what improvement I might see with a CPU upgrade. 

In addition to what you already stated (assuming 40% UTL), please tell me what aircraft you loaded up (C&D or engines running), what 3rd party scenery (if any) for KJFK, and any other scenery add-ons (like Orbx, etc.) in the region.  Also, are dynamic lights enabled and are any turned on in the aircraft.  Thanks for your time.  The 2080Ti produced an 850% performance boost over the GTX1060 I had previously, so I'd be interested to see what marginal improvements might come with a 9900k.

I loaded up the default F22, with the Drzweicki NYC, JFK and EWR installed.  UTL was also installed along with orbx.  on 31L, default spring day, with Fair weather.  Also running ASP4 and rex Skyforce.  Avg was 26-28, now 34-38 on the new card.  Not a scientific test but was good enough considering NY airspace has always been nearly unflyable for me.

[Sartanius 399]

51 minutes ago, pegruder said:

I loaded up the default F22, with the Drzweicki NYC, JFK and EWR installed.  UTL was also installed along with orbx.  on 31L, default spring day, with Fair weather.  Also running ASP4 and rex Skyforce.  Avg was 26-28, now 34-38 on the new card.  Not a scientific test but was good enough considering NY airspace has always been nearly unflyable for me.

Perfect... and thanks, Chris.  Gives me something to go by.  I am currently getting a solid 30 FPS in a similar scenario, but I'm using FSD KJFKv2, FSL A320 (eng running) and all the Orbx products.  Also using P3D "Clear Skies" without ASP4 and ASCA.   Tried my scenario with the default F22 and got a solid 37-38.  I agree... not scientific, but encouraging results.

Merry Christmas!