Nvidia GeForce Game Ready Driver 417.35 WHQL Has Been Released

cageymaru

cageymaru

Fully [H]
Joined
Apr 10, 2003
Messages
22,690
The Nvidia GeForce Game Ready Driver 417.35 WHQL has been released and it adds support for Deep Learning Super Sampling (DLSS) beta in Final Fantasy XV: Windows Edition. The following 3D Vision profiles have been updated: Insurgency Sandstorm has a Not Recommended rating and Atlas has a Fair rating from NVIDIA. The release notes are found here. Special thanks to schlitzbull !

Open Issues in Version 417.35 WHQL. [GeForce GTX 1080Ti]: Random DPC watchdog violation error when using multiple GPUs on motherboards with PLX chips. [2079538]. [Firefox]: Cursor shows brief corruption when hovering on certain links in Firefox. [2107201]. [G-Sync]: Random flickering occurs when connecting G-sync monitor + non-G-Sync HDMI monitor with G-Sync enabled. [2399845].

Fixed Issues in this release include: [SLI][TITAN Xp]: SLI is disabled by default after installing the driver. [200471881]. [TITAN V][NVIDIA Control Panel]: The Workstation->Manage GPU Utilization page appears when it shouldn't. [200470813]. [Rocket League]: The game launches to a white screen with audio in the background and then crashes. [2451530]. [Battlefield V: Day0 97][Ansel]: After being moved all the way to the left, the Ansel field-of view (FoV) slider stops following the click-and-drag mouse movement. [2438857]. [Hitman 2 Silent assassin]: There is flickering texture corruption in the game. [200472315]. [Notebook][3D games]: Frame rate of 3D games may drop to under 30 fps on notebooks. [2456653].
 
"Frame rate of 3D games may drop to under 30 fps on notebooks." - I don't think that's an issue Nvidia is going to be able to fix :p
 
Ooooo DLSS. For a game I don’t care about but maybe we’ll get some good analysis.
 
Will this make half my games black screen like 417.22 did? Also thanks for making me have to manually set buffer flipping mode to on so opengl works.
 
DLSS looks horrible

all dark edges have white borders, the boat the windows all have white jagged edges, the Radars had it, hell the radar platform has it.

Then in the car the Glove box area has really jagged white line through it.

Not subtle effect either, it draws your attention right to it
 
decent I suppose...

4K (DSR), max settings excluding NVIDIA proprietary stuff & motion blur

DLSS - 68-71fps
TAA - 58-62fps
 
They don't even compare the same card to the same card, so we can't see just the effect of DLSS. They compare a 1080 Ti to a 2080 Ti to make the FPS gain look artificially high.

Why not just test a 2080 Ti w/ DLSS vs. 2080 Ti w/ TAA? :whistle:
 
Darkswordz said:
They don't even compare the same card to the same card, so we can't see just the effect of DLSS. They compare a 1080 Ti to a 2080 Ti to make the FPS gain look artificially high.

Why not just test a 2080 Ti w/ DLSS vs. 2080 Ti w/ TAA? :whistle:
Click to expand...

You mean a fair compariton? from nvidia? keep dreaming...
 
Darkswordz said:
They don't even compare the same card to the same card, so we can't see just the effect of DLSS. They compare a 1080 Ti to a 2080 Ti to make the FPS gain look artificially high.

Why not just test a 2080 Ti w/ DLSS vs. 2080 Ti w/ TAA? :whistle:
Click to expand...

Nvidia did the same shit when they announced the RTX cards and people were in a frenzy trying to defend that shit and say it was totally fair. No reason for them to stop doing it now
 
homernoy said:
Haha-ha?
Click to expand...
He meant to add without lighting on fire. There's been issues with cards self destructing. HardOCP's card I believe was one such card. Didn't make it thru the test before deciding to end it's own life.

Darkswordz said:
They don't even compare the same card to the same card, so we can't see just the effect of DLSS. They compare a 1080 Ti to a 2080 Ti to make the FPS gain look artificially high.
Why not just test a 2080 Ti w/ DLSS vs. 2080 Ti w/ TAA? :whistle:
Click to expand...
Probably to market towards people sitting with a 1080 or 1080TI and saying to themselves "$1000 dollars? What the heck!? Why would I upgrade to a 2080TI. The 1080 does everything I want it to. It does 4k "gud enough".

Those people sitting with the card on the left might look at the water reflections accurate showing the buildings in the distance and a few other kind of nice lightning moments....and go meh not worth $1,000 at all....then look at the frame rate and go..."Well, I do have that income tax rebate check coming in, etc". It's mostly trying to convince cheapstakes to get a bit of green-envy when seeing 70 FPS rather than 40-44 FPS. Not that 40-44 FPS is terrible but a steady 60FPS and 4k with high detail is unarguably a better gaming experience. Will it tempt anyone to upgrade if they are sitting at a 1080 or 1080TI? Who knows. But they are trying to sway those people.
 
ZyzzyxSilver said:
They've boosted brightness and saturation on the DLSS side of that video (or dulled the other side) to artificially make DLSS look better. It's particularly obvious in the thumbnail, but it's there the whole way through.
Click to expand...
I think they either used more compression in the video or else blurred the content slightly on the TAA video also. It's kind of impressive how non-subtle they are about trying to rig this here.

Remember kids, run antialiasing for brighter colors and sharper textures because that's exactly what it does!
 
shad0w4life said:
DLSS looks horrible

all dark edges have white borders, the boat the windows all have white jagged edges, the Radars had it, hell the radar platform has it.

Then in the car the Glove box area has really jagged white line through it.

Not subtle effect either, it draws your attention right to it
Click to expand...

Completely agree.

Can't wait for DLSS to get slammed when people see how fugly it is, especially with alpha textures like fences etc.

Saw that coming.

I also like the way they compare it to TAA, which is a smudgy mess akin to FXAA and it still looks worse, lol. :ROFLMAO:
 
schlitzbull said:
decent I suppose...

4K (DSR), max settings excluding NVIDIA proprietary stuff & motion blur

DLSS - 68-71fps
TAA - 58-62fps
Click to expand...

That isn't accurate, as the TAA demonstration side is running on a 1080ti, not a 2080ti. This whole video is a scam. If they want to compair TAA vs DLSS, both have to be shown on the same hardware.
 
For what it's worth, this driver fixed an issue I was having with Nestopia of all things.
 
spine said:
Completely agree.

Can't wait for DLSS to get slammed when people see how fugly it is, especially with alpha textures like fences etc.

Saw that coming.

I also like the way they compare it to TAA, which is a smudgy mess akin to FXAA and it still looks worse, lol. :ROFLMAO:
Click to expand...

Almost every form of AA adds some degree of blur to textures, regardless of what their solution is. If you can play at a really high resolution (even at Super Sampling) on a smaller display, you will get the best visuals possible, otherwise it becomes a tradeoff. This video is misleading as they're comparing a 1080ti to a 2080ti and therefor irrelevant. The whole DLSS scenario is an interesting case where Nvidia has some hardware, that has the ability to do something (tensor), and they have to software what that is exactly, whereas in the past you made hardware for a specific reason. I know it's a stetch but it's almost like local police forces buying straight up heavy military weapons to police small time offending local citizens. You have this extra modicum of power, and feel obligated to substantiate the standardization of that new ability into your repertoire, despite not needing a 50-cal or white phosphorus to direct traffic during a parade.
 
Hmmm... Maybe Nvidia should put all those tensor cores to work actually doing something useful like mining in the background while you game so that it can help defray the absurd cost of purchasing an RTX card in the first place.
 
Well I'm curious...Why not do the comparison on the same hardware? Why is it a 1080ti and a 2080ti? Way I see it, the comparison should have been on the 2080Ti only.
 
alxlwson said:
Well I'm curious...Why not do the comparison on the same hardware? Why is it a 1080ti and a 2080ti? Way I see it, the comparison should have been on the 2080Ti only.
Click to expand...

Simple. The comparison wouldn't have looked as good if it were apples to apples. So they went oranges to bananas instead.
 
You must log in or register to reply here.
Back
Top