NVIDIA Hotfix GeForce 496.61 WHQL driver download & Discussion

Yeah, I am on ivy-bridge 3570k and have MSI enabled on each driver update for the past few yrs, very few crashes overall here (1 BSOD in 2 -3yrs or so).

 

PCIE 1.0 - specifically the Nforce series implementing PCIE 1.0, pascal and newer can't run on gen 1 anyway in my experience, the gpu will post and you can get into windows on the basic driver but driver installation results in black screen.
for PCIE 2.0, Turing doesn't work on some Gen 2 boards that don't have the full ACPI 2.0 tableset implemented (Rampage II for instance, Rampage III does but the default for it is disabled, ie, no post without swapping from an older card to turn it on)

Rev1 x58's(Rampage II) have a MSI errata anyway, the IOH doesn't service MSI requests so they are directed to the ICH instead.

Improvement on floor noise on either end of the bus could bring the product behavior into a more agreeable state, maxwell can do gen 3 on Ivy-E, Pascal can do gen 3 on Sandy-E.

If you want, the LGA 2011 errata covers specifically what the 8Gt/s issues are,
https://www.intel.com/content/dam/w...tes/core-i7-lga-2011-specification-update.pdf

BS55, BS69 and BS75 and BS142 are most relevant . . . . actually there are many more rlevant well beyond BS150, listing them all would be a task.

 

First flickering, now I am getting GSODs...

Memory Dump Output
https://pastebin.com/KGaLL7Jt

@ManuelG

Windows 11 22000.282, GTX 1660
_
The flickers happen usually when a video gets played which matches the dump. This time it didn't just flicker, it shut down my computer and I could see the video on top of the BSOD.

 

Ive never had any NVRLA running?

 

Does it go away if you limit the FPS to say 120FPS?

 

Any news on a new optimized driver for Forza Horizon 5 wish will be available to play in 23 hours ?

 

Also, is this game whitelisted in drivers to use ResizableBar like Forza Horizon 4?

 

496.49 is FH5/Vanguard GRD.

 

Thanks for the heads up , i'm on the latest Beta 496.61 , i guess this is it

 

Why do you always sound like a mad Nvidia stan? Everyone here uses their products and drivers. We all like Nvidia, too... but at least half of us have a ton of issues with their non-Standard versions of these drivers.

It's not like we are making up problems, that would be a complete waste of all of our time, and most of us never had issues with the drivers before August 2020. We'd never really been complaining on the forums a lot, until November 2020 to July 2021.

Turned out, that was due to an actual driver issue, the DPC Latency problem. All the while, you and others tried to say that nothing was wrong with the drivers for those 8 months.


Basically noone had issues with the 471.96 Standard drivers nor the 472.12 Standard drivers, yet a lot of people have had issues with the most recent drivers which have went to the "DCH only" release format.

Is it really more logical to assume that all of these users are "making up problems in their head" or are "too tech illiterate" at the exact same time... for no real reason?

Hear me out, isn't it more logical to assume that perhaps there are actually performance issues and numerous bugs in the most recent drivers, given that you can see exponentially more people complaining similar problems?


All you ever do here is try to tell people that they are wrong, that they do NOT experience what they are all complaining about, and defend the drivers acting as if they are "perfect". You keep it up if you wish, if you really think there's a point trying to tell people that they are making up issues.

I just hope you realize how ridiculous you sound to everyone else, half the time.

 

Yes and I was one of the ones banging on about it for months while Astyanax said it was my PC and wasn't a Nvidia Driver issue.. I knew otherwise

 

I agree. A bug that has been fixed in the standard driver appears in the DCH

 

You can all herp and derp and pat yourselves on the back all you want, but once again: dch and standard core drivers are the same.

 

Detroit BH is a real B* to fix with Nvidia gpus

 

While I can't stand the guy, there's zero difference between the drivers packaged in the GRD and DCH packages. They're exactly the same binaries, installed into exactly the same places, registered into the system exactly the same, etc blah blah. The only difference is that the NV Control Panel and GFE are shipped out of band.

If you think there's a difference, there's none. You're welcome to run checksums over the core driver files, tho, if you think otherwise.

 

Nope, doesn't look like it, unfortunately.

Indeed.

 

An issue with Z390(and maybe later) motherboards and Geforce cards with drivers beyond r451 has been narrowed down to the PEG-ASPM capability, this was brought up on the Asus forum last year but it appears to be Coffee Lake specific and covers the entire Gen 3 range of Geforce cards.

This setting combined with Gen 2 (Card or by configuration) does not demonstrate the issue.

This issue may manifest as the VPU unit listing maxed load at startup, combined with a TDR upon starting hardware decoded content.

 

Uhhhh... I understood Hardware Decoded Content and I have been seeing flickers when watching videos so I guess that's it.

 

I have split my two G70A from all on HDMI (after the driver side fixes) back to HDMI+DP because of some odd crashes, that went away by this action.

Now I have the DP sleep(?) bug mentioned by other people recently. It only happens after long idle. If I go to the crapper for 20 minutes and come back, it wakes up fine. If it idled over night and I want to use the system, the DP display wakes up but stays black, and then goes back to sleep for an ostensible lack of a DP signal.

I'm not sure what's going on, i.e. a driver issue or a display issue (it's a little funky in regards to the very least EDIDs, anyway). The way the displays are hooked up worked fine in regards to sleep with the 472 drivers.

 

Do you have Hibernation/Hybrid Sleep enabled in Windows, by any chance? If so, try to disable those and see if it helps.