The RTX 2080 was first sold in Sep. 20, 2018. Earlier driver before this may not install. One FB forum member is using the Nvidia Creator Ready Drivers. He claims a good increase in Rendering speed. See this article: https://blogs.nvidia.com/blog/2019/03/20/rtx-creator-ready-drivers-supercharge-apps/ . It works for videographers and photographers.

Quote The RTX 2080 was first sold in Sep. 20, 2018. Earlier driver before this may not install. One FB forum member is using the Nvidia Creator Ready Drivers. He claims a good increase in Rendering speed. See this article: https://blogs.nvidia.com/blog/2019/03/20/rtx-creator-ready-drivers-supercharge-apps/ . It works for videographers and photographers.

Thank you Tomas, this sounds good. I'll give it a try. Wish me luck.

I think I'm screwed

The latest Nvidia driver is 430.39 WHQL. I do not understand the foreign language for the 419.67 screenshot you posted here. I believe that certain builds of windows 10 can only accept only certain Nvidia drivers. You can look up this information on the internet or hope that someone that is more familiar with win 10 can come out with a workaround for this.

I believe that all you have to do is to uninstall your present nvidia driver first and then do a custom install to only install the basic driver and nothing else. That should work so you can do your desired tests.

Please take a look at the Read Me Before Posting sticky thread and attach your DxDiag test results so we can see the specific version details of your system. You may also have unintentionally downloaded a Win7 driver version from the nVidia site, or a 64-bit version but you have a 32-bit Windows.

Like tomasc mentioned, try deleting the existing driver so Win10 will replace it with the basic VGA driver. When you download 430.39 for your exact version of Win10, use these settings:







It's ok to install the USBC driver too, but you want all the other boxes unchecked. If you're a gamer, go ahead and install the PhysX software as well.

YouTube/optodata

DS365 | Win11 Pro | Ryzen 9 3950X | RTX 4070 Ti | 32GB RAM | 10TB SSDs | 5K+4K HDR monitors

Canon Vixia GX10 (4K 60p) | HF G30 (HD 60p) | Yi Action+ 4K | 360Fly 4K 360°

Quote I do not understand the foreign language for the 419.67 screenshot you posted here.

Only the first sentence is Dutch. The rest is English :

I believe that all you have to do is to uninstall your present nvidia driver first and then do a custom install to only install the basic driver and nothing else. That should work so you can do your desired tests.

I tried that - I deleted all drivers and even all references to Nvidia in the registry. I did that because the install log from the failed installation said

50.708 | INFO: [PackageUtil] 682@PackageUtil::FetchDriverTypeUsingRegistry : Detectoid Registry value found, driverType is DCH.
50.708 | INFO: [PackageUtil] 616@PackageUtil::FetchDriverType : Installed driver is DCH.
53.371 | INFO: [PackageUtil] 539@PackageUtil::IsToBeInstalledDriverUniversal : The most compatible driver for "PCI\VEN_10DE&DEV_1E82&SUBSYS_37221462&REV_A1\4&1C3D25BB&0&0019" is "c:\nvidia\displaydriver\419.67\win10_64\international\display.driver\nv_dispci.inf".
53.529 | INFO: [PackageUtil] 545@PackageUtil::IsToBeInstalledDriverUniversal : The driver to be installed is not a DCH driver.
53.529 | INFO: [PackageUtil] 667@PackageUtil::FetchDriverType : Selected driver is 1.
53.529 | ERROR: [PackageUtil] 591@PackageUtil:ecideBlockInstall : Blocking installation for driver compatiblity.
53.529 | ERROR: [DisplayDriver.DisplayDriverExtSite] 1181@CDisplayDriverExtSite::CheckConstraint : Blocking the install as the already installed driver and to be installed driver are not compatible.
53.530 | ERROR: [NVI2.Installation] 1046@CInstallationCheckingBase::CheckConstraints : Critical package blocked, turning to error "BlockInstallForCompatibility".

...so I thought that cleaning up the registry might do the trick. Unfortunately it didn't. However, I read that the awful GeForce Experience software was able to install Creators drivers so I reluctantly installed it and used that to install the Creators drivers. So far however it has not changed anything for the speed of the editing. I don't see any difference.

Quote Please take a look at the Read Me Before Posting sticky thread and attach your DxDiag test results so we can see the specific version details of your system. You may also have unintentionally downloaded a Win7 driver version from the nVidia site, or a 64-bit version but you have a 32-bit Windows.

Like tomasc mentioned, try deleting the existing driver so Win10 will replace it with the basic VGA driver. When you download 430.39 for your exact version of Win10, use these settings:

I don't even get to see this screen when I install the Creators driver. I get an error and that's it. But I managed to install the Experience software and was able to install the Creators drivers.

I'm glad you got things working again!

As I wrote in your other thread about frustrating titles performance, I didn't think that the problem was related to your nVidia card or driver, and you've now proven that convincingly. Still, this is progress in eliminating possible suspects

I'm downloading your shared project now and will post soon on the other thread.

YouTube/optodata

DS365 | Win11 Pro | Ryzen 9 3950X | RTX 4070 Ti | 32GB RAM | 10TB SSDs | 5K+4K HDR monitors

Canon Vixia GX10 (4K 60p) | HF G30 (HD 60p) | Yi Action+ 4K | 360Fly 4K 360°

Quote The RTX 2080 was first sold in Sep. 20, 2018. Earlier driver before this may not install. One FB forum member is using the Nvidia Creator Ready Drivers. He claims a good increase in Rendering speed. See this article: https://blogs.nvidia.com/blog/2019/03/20/rtx-creator-ready-drivers-supercharge-apps/ . It works for videographers and photographers.

Lots of good articles out there and lots of videographer tools (free and purchased) that utilize the improvements Nvidia put into the new NVENC, SDK and the Turing GPU. Unfortunately, none of this really relevant for PD regardless what driver you utilize until CL provides a PD version which utilizes the new Nvidia API. So, until CL announces support, one can only think how things might be with PD someday or evaluate capability in other tools currently.

Jeff

Quote Unfortunately, none of this really relevant for PD regardless what driver you utilize until CL provides a PD version which utilizes the new Nvidia API. So, until CL announces support, one can only think how things might be with PD someday or evaluate capability in other tools currently.

In the forums there are claims that seem far fetch but a church goer claimed replacing his geforce driver to the creators driver reduced his previous 1 hour rendering time to 8 minutes. That is a 7.5x decrease in rendering time in PD17. It should help those who may have a slow pc and want faster rendering. See the attached.

Nvidia card users may want to try the creators drivers with PD17 and report the speed increase they get in rendering or if it makes editing faster. More than one user has endorsed this.

Quote In the forums there are claims that seem far fetch but a church goer claimed replacing his geforce driver to the creators driver reduced his previous 1 hour rendering time to 8 minutes. That is a 7.5x decrease in rendering time in PD17. It should help those who may have a slow pc and want faster rendering.

I'll bet my soul that 7.5x decrease was not from features exposed in a driver change. Not doubting that this church goer experienced it, but just not the effect of the drivers alone, 100% impossible. What this sounds more like possibly hardware encoding was not working correctly with the 1050ti and then after a driver change it started to work. Probably more of a configuration issue that gave the 7.5x observed.

Jeff

I just did some tests to see if I could find a noticiable difference between different nVidia drivers.

I don't know any details of which versions the poster in the other thread was using, or what kinds of clips or edits were in their projects. My test project is 5:35 long and it's a mix of H.264 60p 4K and HD clips with titles and lots of Crop/Zoom/Pan edits.

Here are my results from 3 recent driver versions when using Fast video rendering technology:

Output / Driver	419.67 Game Ready	419.67 Creator Ready	430.64 Game Ready
HD 60p AVC M2TS	8:30	8:40	8:21
4K 60p AVC M2TS	31:52	33:01	31:55
HD 60p HEVC M2TS	8:15	8:46	8:21
4K 60p HEVC M2TS	33:42	35:19	35:12

Nothing here indicates that one version was significantly better than the others. If anything, it looks like the Creator version is a little slower with my test project, and I don't see any reason to not keep using the latest 430.64 GR driver on my system.

YouTube/optodata

DS365 | Win11 Pro | Ryzen 9 3950X | RTX 4070 Ti | 32GB RAM | 10TB SSDs | 5K+4K HDR monitors

Canon Vixia GX10 (4K 60p) | HF G30 (HD 60p) | Yi Action+ 4K | 360Fly 4K 360°

Excellent test Optodata! Thanks for all this work.
I wonder what the timing would be without the Nvidia encoding : Would having and using Nvidia really make a huge difference compared to Intel?

Optodata – Thank you for taking the time to check out the Nvidia creator driver. Your test above has help to debunk the creator drivers use for PowerDorector.

Quote Optodata – Thank you for taking the time to check out the Nvidia creator driver. Your test above has help to debunk the creator drivers use for PowerDorector.

I can confirm that. There is no difference in performance between the game or creator driver. But then again, no option in PD17 makes a difference in my case - switching OpenGL on/off, Hardware acc. on/off - when it comes to editing & the preview screen it doesn't change a bit.

Quote Excellent test Optodata! Thanks for all this work.
I wonder what the timing would be without the Nvidia encoding : Would having and using Nvidia really make a huge difference compared to Intel?

Well, I took your bait and did the same test with my UHD Graphics 630 GPU:

Output / Driver	26.20.100.6709
HD 60p AVC M2TS	9:55
4K 60p AVC M2TS	37:01
HD 60p HEVC M2TS	10:06
4K 60p HEVC M2TS	39:44

In the interest of full disclosure, I didn't run these 16 production runs all the way through; For the HD versions, I let PD run for exactly 60 sec and added 1 min to the estimated remaining time. For the 4K versions, I ran them for 10:00 and added 10 min to the remaining time.

Because of the complex nature of the test project, the actual production times might be different than the numbers I've reported, but if there were a difference of anything like 7-8x (or even down to a 40-50% improvement) it should have been apparent in these tests.

Quote

Well, I took your bait and did the same test with my UHD Graphics 630 GPU:

Output / Driver	26.20.100.6709
HD 60p AVC M2TS	9:55
4K 60p AVC M2TS	37:01
HD 60p HEVC M2TS	10:06
4K 60p HEVC M2TS	39:44

In the interest of full disclosure, I didn't run these 16 production runs all the way through; For the HD versions, I let PD run for exactly 60 sec and added 1 min to the estimated remaining time. For the 4K versions, I ran them for 10:00 and added 10 min to the remaining time.

Because of the complex nature of the test project, the actual production times might be different than the numbers I've reported, but if there were a difference of anything like 7-8x (or even down to a 40-50% improvement) it should have been apparent in these tests.

This is excellent again Optodata! Thanks!
Good to know of your testing approach. I fully agree with your last statement. This test (and we should keep the results somewhere at hand) shows that in general, using NVidia gives some, but not a huge advantage in production time. We assume that the output quality is the same for each, but we save that maybe when there is a long and cold winter.
Thanks Again!

Quote
Good to know of your testing approach. I fully agree with your last statement. This test (and we should keep the results somewhere at hand) shows that in general, using NVidia gives some, but not a huge advantage in production time.

Warry, That potentially is a very misleading statement. The encode performance of a particular project is very much tied to the features incorporated in that project and hence not a good reflection on various encoding technologies as standalones.
Like the test that was here: https://forum.cyberlink.com/forum/posts/list/78189.page#post_box_320609
of similar comparison goal but everything was CPU bound because of the edits so it showed little other than my 4yr box with a basic GTX1070 was 50% faster than a new box built specifically for video editing. 50% is a rather larger difference.

Unknow what this project entails other than:

Quote My test project is 5:35 long and it's a mix of H.264 60p 4K and HD clips with titles and lots of Crop/Zoom/Pan edits.

and

Quote Because of the complex nature of the test project,

If you really want to compare encode technologies, one needs a basic transcode, not a full edited project, just too many more variables potentially at play.

Jeff

Quote If you really want to compare encode technologies, one needs a basic transcode, not a full edited project, just too many more variables potentially at play.

Jeff

No one is stopping you from doing a full test. I look forward to your results!