Nvidia GeForce 364.51 beta WHQL driver download & discussion

I think you should stop with these stupid threats to an NV's representative who obviously can't give you an answer. Feel free to proceed with the law or whatever but don't turn this thread into your own personal battle. I think you know that you will only get an answer when NV will be ready to give it to you, not earlier.

 

Maybe I came across a bit harsh, sorry.

I'm just thinking of how things went when dx11 came out. Almost all d3d11 functionality could be emulated in d3d9, but it would not be hardware accelerated. Granted, dx12 seems to involve more how windows handles the workload between cpu/gpu, so i could completely be off on this. They probably had a harder time than expected when they went to actually implement dx12 in Fermi?

 

I am still on 364.47 WHQL driver and i have no SLI / DP issues except some flickering in Fl-studio12
so it is worth upgrading to this one?

 

No stupid threats here. This is the job of nv rep to do. If he/she can't do it, better to find different job. We asked him very politely and he refused to give any information about the state of the driver. This forum is to give question to them, so i just use my free rights to do it. thousands and thousands of Fermi owners are angry.

 

I wasn't having any issues either but I changed to this driver just to be safe and seeing as Nv pulled it that solidified my decision.

 

The certified non-beta version of these drivers are just as bad for me as the previous two sets. Came home again to a shutdown computer. Powered on to watch DirecTV online streaming and it froze 5 minutes later with a solid color on one display and black screen on the second.

This is after playing Divinity: OS, The Division and ROTTR for a combined 11 hours the past day. Can't idle or do low power tasks on any 364 driverset with my setup it seems.

 

i think i have this too. before it was 1.4-1.7GB RAM in use with nothing running and now it's 2-2.3GB. :3eyes:

 

They're ok for me, no basic desktop issues or anything, I even flashed my gpu to a little higher boost a while ago @ 361.82 (stock 1342 > now 1404mhz, vram 7020 > now 7240mhz) and still uber fine.


I just want a proper dx12 boost soon, something like it happened by DX11 and R337.50 driver, more index, vertex buffers, more driver calls etc.


The sooner the better.

Common nv do some magic, its been a while. :nerd:

 

Would it be possible to download and install the latest Vulkan API manually and/or is it driver bound?

 

Installed this and got two PC freezes in two days. First time my PC froze with a repeating sound, and just not both my monitors went black and I had to restart manually.

Downgraded to 361.91. Anyone tried DX12 TR with the old one?

 

mine 2.5-3.1 GB eventhou, i got 8 GB ram only.

 

Edit

 

here is my API overhead results of 3DMark compared to an older driver

Left is the new driver

http://www.3dmark.com/compare/aot/123576/aot/116205

That Is a lot of green I must say

 

I think it's driver-bound at least for now.

https://github.com/KhronosGroup/Vulkan-Docs
https://www.khronos.org/vulkan/

 

It will always be driver bound as any other API ever was.

 

Drivers seem to be working fine here.

Just out of curiosity, what are people expecting to get out DX12 on their Fermi based cards? They reached their limits a long time ago...

 

It's just a matter of principles and I for one happen to share their concerns.

 

DX12 doesn't automatically mean the games will be super demanding.

The more cards that support DX12, and even more so Vulkan though (due to not being limited to Windows 10), the faster the studios can drop their DX11 support.
The main reason to keep using the DX11 API would then be for undemanding games that would work on DX9-11 GPUs.

400/500 series with DX12 and Vulkan support would mean all Nvidia's DX11 and OpenGL 4 cards could be targeted by just using either DX12 or Vulkan.

The main complaint however, is that Fermi DX12 support was promised two years ago.
I do think the majority of Fermi users would appreciate an open statement from Nvidia regarding the DX12 status, so Fermi users won't feel left in the dark. Even if Nvidia says it won't be here until summer, that'd be better than simply ignoring all questions regarding Fermi.

Perhaps Nvidia underestimated the resources required to bring DX12 to Fermi, and perhaps Nvidia overestimated Fermi's capability to run the games of 2016, but backtracking on the promise now would be a severe blow for Nvidia's PR.

 

FERMI:

it's better to burn out than fade away . . .

 

DX12 is a big investment development wise and it's very unlikely that you'll see many non demanding indie games using it in the following years. Most AAA games which will use DX12 won't run well on Fermi GPUs in any API.