bruh, what? you joking, right? all of these new drivers are so bad lmao
Could be system and game dependent, on my aging system ( 1060 6GB / Intel 2600K / Z68 / X-Fi Titanium PCIe / 16GB DDR3 1800 ) the 511.xx drivers meet or exceed the 'golden drivers' I've used in the past for the games played.
Summary of tests from 442.49 to 511.65, trying to test the same games as Lokkenjp over on Reddit and generally match. Full breakdown with data in Spoiler, a summary follows after
[SPOILER="All Data"]
https://i.imgur.com/pHszT1d.png
[/SPOILER]
Summary, highest performing driver for each game and category ( FPS, frametimes AVG 1% 0.1% )
Note that the spreadsheet 'summary' column automatically selects the 'best' driver for each game in each of the 4 categories ( FPS, frametimes AVG 1% 0.1% ) .If a category result 'matches' across multiple drivers for a given category the higher driver number in the result is selected e.g. if 442.49 and 511.23 have matching FPS results then 511.23 is selected in the 'summary' list.
EDIT 1: Summary with 511.xx driver test results removed so spreadsheet selects drivers from 442.49 to 497.29
Compared to the previous summary the spread is quite large, the 511.xx series are predominant in previous summary.
EDIT 2: Added full data in Spoiler 'All Data' tag
This driver is good as previous, but without gsync issue.
I like it, has got my best 3Dmark Port Royal result in my setup, https://www.3dmark.com/pr/1432398
Like all the 5xx drivers I've tested, dwm.exe is spamming NV service REG queries ~1000 times a second with these installed. Seems to be just repeating the same pattern for no obvious reason. I don't see any benefit from this branch, so back to 472.47 STUDIO (behaviour is completely absent).
Like all the 5xx drivers I've tested, dwm.exe is spamming NV service REG queries ~1000 times a second with these installed. Seems to be just repeating the same pattern for no obvious reason. I don't see any benefit from this branch, so back to 472.47 STUDIO (behaviour is completely absent).
Did you look at the screenshot I posted from Process Monitor. That's idling at the desktop. If I could be bothered, I'd post another one with 472.47 installed, where the behaviour is absent. Timestamps don't lie.
Please explain, what do you mean by "no it isn't".
Did you look at the screenshot I posted from Process Monitor. That's idling at the desktop. If I could be bothered, I'd post another one with 472.47 installed, where the behaviour is absent. Timestamps don't lie.
Please explain, what do you mean by "no it isn't".
Stop using "slimmed" drivers, things break when you cluelessly remove libraries from the driver
Managed to get some lines for that location by setting the desktop to a DSR resolution,
it is not spamming it at all.
Stop using "slimmed" drivers, things break when you cluelessly remove libraries from the driver
Stop saying stupid things, core driver isnt touched at all when using slimmer, always using slimmer to get rid of garbage, never had problem until now 🙂
Stop saying stupid things, core driver isnt touched at all when using slimmer, always using slimmer to get rid of garbage, never had problem until now 🙂
No comments from the ignorant peanut gallery, circular dependencies have always existed.
Ohmer:
Did you look at the screenshot I posted from Process Monitor. That's idling at the desktop. If I could be bothered, I'd post another one with 472.47 installed, where the behaviour is absent. Timestamps don't lie.
Please explain, what do you mean by "no it isn't".
Your Monitor data is invalid in the registry, delete with CRU, and remove the instances from device manager, let it redetect and reconfigure with CRU as desired, this could be a result of both CRU or Custom DSR tool.
88 instances each for changing to a DSR resolution and back to native.
https://cdn.discordapp.com/attachments/247515315825672203/939520871394050089/unknown.png
No comments from the ignorant peanut gallery, circular dependencies have always existed.
Your Monitor data is invalid in the registry, delete with CRU, and remove the instances from device manager, let it redetect and reconfigure with CRU as desired, this could be a result of both CRU or Custom DSR tool.
88 instances each for changing to a DSR resolution and back to native.
https://cdn.discordapp.com/attachments/247515315825672203/939520871394050089/unknown.png
How much are you paid to defend Nvidia ? Nvidia Slimmer from uKer working flawlessly, no problem at all and no garbage, maybe the ignorant is you, as you even dont try it by yourself 🙂
How much are you paid to defend Nvidia ? Nvidia Slimmer from uKer working flawlessly, no problem at all and no garbage, maybe the ignorant is you, as you even dont try it by yourself 🙂
Maybe attempt to make this coherent so we can continue to be disappointed in you for clearer reasons.
No comments from the ignorant peanut gallery, circular dependencies have always existed.
Your Monitor data is invalid in the registry, delete with CRU, and remove the instances from device manager, let it redetect and reconfigure with CRU as desired, this could be a result of both CRU or Custom DSR tool.
88 instances each for changing to a DSR resolution and back to native.
In the interests of science, did a full DDU with monitor removal specified (as usual), checked Device Mgr on reboot, gone.
Re-installed the full package ("got a clue" in your parlance) for 511.72 with only GFE deselected at the installation screen. LAN re-connected, rebooted, idle at desktop, ran Proc Mon with the same filter as you. Same result ~1000 entries a second.
Monitor appears to be correctly detected in both Device manager and NVCP. I'm not running DSR, or any custom resolution. My guru3D profile is up to date. I have an Ultrawide LG (38WK95C-W UW FreeSync) running at native resolution of 3840 x1600. Never had a problem with it.
It's late here, so I'll have another look tomorrow, but I know if I re-install 472.47, the issue will vanish. I've never seen it before I started testing the 5xx series, and I use Proc Mon quite frequently, especially after adding new hardware/drivers, and not just for video stuff.
Framerate(FPS) 511.65 Division2 456.71 Farcry5 511.65 ForzaH4 511.23 Wildlands 456.71 WorldTanks FrametimesAVG 511.65 Division2 511.65 Farcry5 511.65 ForzaH4 511.65 Wildlands 511.65 WorldTanks Frametimes1% 497.09 Division2 511.23 Farcry5 511.65 ForzaH4 511.65 Wildlands 511.65 WorldTanks Frametimes0.1% 511.65 Division2 511.65 Farcry5 442.59 ForzaH4 496.49 Wildlands 511.65 WorldTanks
Note that the spreadsheet 'summary' column automatically selects the 'best' driver for each game in each of the 4 categories ( FPS, frametimes AVG 1% 0.1% ) .If a category result 'matches' across multiple drivers for a given category the higher driver number in the result is selected e.g. if 442.49 and 511.23 have matching FPS results then 511.23 is selected in the 'summary' list. EDIT 1: Summary with 511.xx driver test results removed so spreadsheet selects drivers from 442.49 to 497.29Framerate(FPS) 497.09 Division2 456.71 Farcry5 496.76 ForzaH4 466.11 Wildlands 456.71 WorldTanks FrametimesAVG 497.09 Division2 456.71 Farcry5 496.76 ForzaH4 466.11 Wildlands 466.11 WorldTanks Frametimes1% 497.09 Division2 456.71 Farcry5 442.59 ForzaH4 466.11 Wildlands 497.29 WorldTanks Frametimes0.1% 497.09 Division2 497.09 Farcry5 442.59 ForzaH4 496.49 Wildlands 496.13 WorldTanks
Compared to the previous summary the spread is quite large, the 511.xx series are predominant in previous summary. EDIT 2: Added full data in Spoiler 'All Data' tag