Discussion in 'Videocards - NVIDIA GeForce Drivers Section' started by Prophet, Nov 21, 2021.
That is absolutely irrelevant to the discussion.
Yeah, but I was only responding to another post that had already brought it up.
You can get these DCH drivers without the store now anyway so its not a big deal anyway.
After I installed drivers CP was not present when I right clicked on desktop. I checked for Nvidia in the start menu and there was the Nvidia CP. I opened the CP and was required to accept an EULA.
In all my years installing Nvidia driver this is the first time I encountered a separate EULA for the CP.
I deleted windows store and some useless things.
dch version "nvidia control panel" dont work. (496.76)
standart versin "nvidia control panel" work. (472.12)
dch (496.76) + standart (472.12) NVSlimmer_v0.12 "install nvcpl" and "nvidia control panel" work.
So you have to click one button. May as well shake your angry fist at the sky.
just some facts.
if you have MS store turned off, and hence no need to have the appx service running, then no, it does not install the nvcontrol panel.
To have nvcontrol panale install, one must also have appx service running.
The work around is to do the install without ms store enabled and have appx service off. then, Unpack the nvcontrol panel from the driver (7z) and copy it into "nvidia control panel" in nvcorporation in the programs directory.
But without appx service running, no, it does not install.
Lastly, any reasonable person would agree that the standard driver (non dch) is free from any of these issues and that the dch add layers of complexity/additional dependencies.. Are those extra layers of complexity yielding a better performance product? No, there is no performance increase benefit from dch.
it just MS trying to reinvent the wheel, and wanting more control over things, this just my opinion though.
Personal i dont see the point of DCH other then MS want everything going threw store and everyone using the new way to install stuff.
As of drivers being DCH no only i dont think care either way I might not use MS but i dont remove it appx for that matter.
It's an internet forum, not your elementary school hallway.
Bloat is a myth?
Bloat doesn't just include random processes in the background. Bloat includes registry issues (which accumulate over time), processes that seem to be 0% but actually update or poll, many of them take up memory for little to zero reason. Bloat includes software that seems innocuous but actively pulls and polls from online sources (Windows functions often come to mind). The difference between bloatware in a prebuilt and a clean Windows install, especially a stripped down one, is objective and has been tested thoroughly. This is such a strange take from someone with apparently a lot of reputation and knowledge.
One of the most common pieces of advice is to uninstall and remove all unneeded applications to improve performance.
No, they may not be contributing to DPC latency when you're sitting there staring at latencymon, but why is the EA background service currently polling and taking CPU cycles, even if just at a fraction of a percent, when I haven't opened it and haven't even opened it since rebooting? My razer lighting engine helper is taking CPU when I have no lights on, and it should be at 0.0%? Why are so many of these background apps taking memory, even small amounts, contributing to having "significant memory usage" (>10%) just by passively having a bunch of programs not even active or turned on? This awful razer software is using more cpu, even if it is small, than my browser is.
You seem to have the privilege of knowledge and aware of what to avoid already, and consider these things 'myths' because you don't get what "bloat" means for the average user, or are a DPC minimalist, or something, and don't experience these issues an average user might have, and so declare it a "myth".
DPC is not the be all end all of performance and something hypochondriac optimizers point to to prove every point. Oh, look, the EA background process is still using cpu cycles every so often. Gee. I wonder what it's doing. I wonder what all those other programs I steer clear of would be doing if they were allowed to have services on with impunity. The Razer lighting engine helper is using a non-insignificant amount of CPU in order to make sure my lights are turned off.
DPC latency is not always correlated directly to Windows performance itself, or the performance of the software being tested. It was literally used to test audio smoothness in older times when that would actually be a huge issue.
No one is saying, typically, a small amount of avg DPC latency increase is some deal breaker. It's the stutter and constant polling, registry bloat, and disk bloat that often can't be solved by just defragging. Responsiveness of Windows is just objectively better on a clean install.
What a weird take. I find most of the default stuff on Windows to be bloat to begin with, and the OS runs better without them, easily, and it is not hard to see a responsiveness difference.
In regards to the different drivers and their bloat, GFE is bloatware because it will do all of the above. Most of the files/folders inside any given nvidia driver are completely useless and many are user metrics and reporting, ie, things that poll and report.
The fact that you can install a driver and have the same performance with three folders vs 12 should tell you something. Nvslimmer exists for a reason and is not just for anti-telemetry, as much as that is a key point.
I personally had to go find my own driver for my Xbox One controller because the firmware and given ones were absolutely useless and DID NOT WORK. The entire system is a mess because it is poorly coded and poorly maintained. The fact that I had to go dig up a forum thread and have some random person give me a magical .dll file that suddenly let the firmware function as it should have, completely independent of MS Store or updating it, is proof enough to me it's neither needed or beneficial. It was the actual cause of my problems there, because I assumed it would be as you said.
Those results are actually about the bluetooth connection and not the official dongle. ie, a workaround that is not intended for use and is clearly broken. My xbox one controller has zero of these issues because I use the actual wireless dongle for it, and not some bluetooth onboard connection.
It has nothing to do with the drivers.
I agree with you. I don't understand why people can't just replace that word with 'inane'. It's even shorter to write. 'Ridiculous'. 'Absurd'. 'Silly'. 'Hyperbolic'. There are countless other words, but we jump to the most hurtful one to many people. It takes less energy and less effort to be marginally respectful. I don't believe in 'habits are hard to break' because I've broken all of mine from being a kid, and it takes hardly any effort at all if you care in the least bit.
I think the staff member was not doing it in bad faith but just replying to the post and clarifying. It is indeed an internet forum, though, and you should manage your expectations. Most people here will not change or care. They are self-centered and cannot understand how their actions might affect other people because they do not have such a problem, and so it seems 'harmless' to them.
This is psychology though and well beyond the point of this thread.
As a final note: if anyone really needs NVCP but wants these drivers for some reason as well, or if in the future this becomes a problem, maybe consider just learning how to use nvidia Profile Inspector (carefully, because it is a power user program) which does all the same things and more, but in a less pretty manner.
The registry is a memory mapped database, age of install doesn't affect performance period, just search performance.
No application "searches" for its registry information, with windows 10 you would have noticed you can now enter the direct hkey path into the bar at the top, applications do this themselves and the opening of a key directly is always fast
Advice which needs to die.
You're correct, however the issues have occurred for a couple of games when wired or using the official dongle, they are just fewer and farer between
I believe Halo MCC was one case, and another was a more recent AAA title though the name eludes me right now.
I have never known an official microsoft manufactured controller to have issues with its own driver, i know the logitech F controllers have this weird issue where you need to replug them occasionally because on start they are code 41 (the control itself doesn't implement the common class standby states).
Things are a bit different with in wireless mode, particularly with the official dongles since they use several different revisions of mediatek wireless chips and multiple different drivers across the different device ID's, the most recent revision uses the newer netadaptercx driver design instead of the legacy ndis adapter and reprogramming it with a firmware blob.
I found that the official Xbox 360 Controller for Windows drivers newer than around 2009 (which includes the latest drivers that Windows 10/11 installs) can lead to the system being unable to go idle and thus the screen saver not being able to activate and/or the system being unable to shut off the display, HDDs or go to sleep. The workaround is to use old drivers from ca. 2009 (v188.8.131.529 dated 13.08.2009 seems to work fine, for example) or disconnect the controller.
I haven't encountered any such idling issues with drivers for Xbox One or Series X Controllers, though.
This one i've found is connected to steam or gfe being open, but its not just the 360 driver, the same can happen with my XBCD driver which turns original xbox and wired 360 controllers into classic remappable directinput controls, i could never identify a solution apart from exiting steam or disconnecting the control because it happened even with the steam input support disabled, I remember both you and @tsunami231 searching for a solution to it for a while and landing on the driver version downgrade.
The problem is inherently hardware decay and noise (and the device manager power settings not working), applications registering usually filtered out jitter as movement, theres slight differences with the old driver and later drivers where applications that don't use xinput were seeing the jitter and activating whatever keep alive code they have.
This slight jitter is filtered out by xinput as the common class driver has a 10 and 15% deadzone for the left and right sticks.
calibration won't fix it because its settings aren't taken into account by raw/win32 access to the controller.
I am implying that life must be difficult for you as an apparently sentient YouTube comment.
Spoiler: not part of topic just reply to being bought up
yah to this day i still can not use new driver then2009 drivers, and to this day ifyoo still cant figure out why only 2009 works correctly after 20h1 builds, but take that with grain of salt. this issue happen in both d input and x input if there drift happen steam seem to do better job of handle drift? cause i had to apply 25 anti drift to both left and right annoying cause deadzone was to big and i have no drift there the dead zone in steam is probably 3~5% now in steam .but when i was having this issue it was happening regardless if steam was open or close, there is huge dead zone on both left/right analogs for diagonal input which i cant remove though. which happen in dual sense and ds4 when use on my pc. which i think i just somthing funky with how windows deals with contolers and deadzones
as for people argue in hear and call out mods for not doing there job, keep doing so and you will think other wise.
This whole thread is from 1995. Thanks for the laughs.
My theory (and testing seems to confirm it so far) is that the right stick raw data is reporting constant movement that should usually be filtered by the driver, the xbcd developers never actually did this so you can actually see the stick flitting side to side near constantly on xbox 360 controllers, i don't know why the left stick was better quality / less twitchy on most of them but i can even see this crap on the classic Xbox controller too so it seems like an actual design choice or possibly circuit noise at blame.
this is picked up by any application that supports rawinput (which i believe windows.gaming.input is actually layered on) so UWP apps, Steam, GFE, anything that takes gamepad input and bypasses Xinput see's constant movement that should be filtered out normally.
Funnily, 1903 introduced similiar issues on the One pad.
The point is simple, all things in performance being equal, why have the added complexities, services, of the DCH vs the self contained, more complete, standard driver?
Simply, there's no carrot on DCH install over standard
Spoiler: again not part dch topic
it weird bug i have never seen in my Logitec dual rumblepad 2 or the F710 the F710 had own issue which why i went bought the ifyoo gamepad. which worked fine with new drivers till 20h1 build. I have not tried to change back to new drivers since maybe at some point i will. at this point i happy it works and let monitor go to sleep, now if only i could get steam or windows for that mater to remove dead zone from diagonals on left and right stick cause the dead zone which worse then deadzone for up/down/left/right seem to be there even with ds4 and dual sense controllers when use them on pc ,
the issue is standard vs DCH. Having complete standard driver or DSCH does not make the standard driver ancient. The point is rooted in the why and the necessity of DCH since it offers no carrot over the standard driver. As you just make comments without addressing the actual question, you have provided the correct response, thank you.
Since there isn't a performance advantage, nor a convenience advantage, the point is moot.
I really really think I like the new DCH task bar icon better than the old standard one.
But could somebody please show me the old one again?