Have you encountered issues with recent nvidia drivers?

You do right about that, Physx is very poorly optimized at those titles, BLTPS works better than BL2, but with physx my fps was between 60-72, with the driver issues, the gpu goes down to 10% and fps is around 15.

 

I been free of issues since the 290 branch.

Used to have TDR issues before that, but no more of those.

 

If the 2gb 960 is the first maxwell card to get pulled out of the market then there is a good chance it´ll be the first maxwell to stop seeing support.

 

Well the 750ti, 750 have been out quite a long time and they are still supported. Where is it stated by nvidia that the 960 2gb is getting discontinued soon?

 

Every report seems to eventually source back to those completely trustworthy guys at wccf

 

How exactly do you think support for a 2GB of VRAM differ from support for 4GB of VRAM on the same chip? They compress some textures on the fly to fit them in 2 GB or what?

Nothing will happen to a 960 2GB even after all Maxwell cards will be deprecated and switched to Pascal line.

Also - this is the first time I've heard that 960 2GB is discontinued.

 

Maybe thats for the better. Having a more Vram this days isnt a bad thing. Availability of R9 380 4GB is quite higher than 2GB, maybe nvidia responded to that. Pricing is also a joke, there is 20-25€ difference between 2GB and 4GB versions.

 

I doubt a 960 can effectively use 4gb of vram with a 128 bit bus. A 960 is exactly 1/2 of a 980 spec wise. Not exactly a powerhouse to use double the vram.

 

Polls that require certain requirements should always be public. Maybe a mod can turn it public.

 

The only bugs I've noticed are minor hitching in games and G-sync would sometimes crash the system when resuming from sleep after extended periods. Other than those two, everything else seems fine. Haven't had time to extensively test the 358.50's yet, but so far G-sync isn't crashing the system after sleep/stand-by. Not sure about gaming yet.

 

It´s not rare for games to need memory management fixes via driver, if some game works bad on 2gb but well on 4gb because it has room to spare, the fix for the 2gb model might come only after several months... because eol cards are not nvidia´s priority. And on the developer side, they wont be able to get their hands on 2gb 960´s to test anymore.

 

It's actually very rare and such fixes are working on all VRAM sizes as they simply optimize some fixed part of memory consumption. There are no size specific optimizations as this is rather pointless if you consider the amount of VRAM configurations which NV driver have to support.

Also even if 960 2GB is EOL there certainly are other 2GB cards which aren't. Stop spreading FUD.

 

In witcher 3 my dsr scaling is broken. I have a full HD TFT and in higher resolutions i cant see the complete screen and in lower resolutionen it doesnt fill up the screen size. In game options i selected scale to full screen. Dont know if its a game or a driver bug. (WIn 10 X64 with a GTX 970)

 

There is always going to be an issue of one sort or another. The driver software is huge. It's never going to be perfect. And hopefully most major issues will get resolved eventually without breaking anything else.

 

Was just watching his last video. DudeRandom84 testing TW3 1.10patch, 980ti. Its a stutter fest for some odd reason.

https://youtu.be/qfsFFde6kyo?t=102

 

Prerender limit set to 1 should solve this. This is an application issue, not the driver's.

 

I can prob post more amd driver problems in the amd section if I wanted, but whats the point in that? To flame or stir the pot?

 

Well, finding the solution, like dr_rus pointed out. Now we know more.

 

not happening here with a gtx970. i'd say it's a recording issue with shadowplay and driver set used, or an issue with his rig.

i have the same OS, 358.50 driver, and a 970. DSR works fine, so it's an issue specific to your monitor/rig.

 

I've been having an issue where anything after 355.84 causes my pc to restart after waking up the monitors(pc does not go to sleep, just my screens).
It seems to be duration based, where if they go to sleep and I come back soon they're fine. But every morning when I wake my PC, it restarts.
Win7 x64. I've tried every WHQL release since 355.84 dev release with the same result. Always end up going back to 355.84 because it doesn't exhibit this problem.