NVidia driver 337.50 beta released!

[krumme]

It's not often you see 10% improvement with single cards (Not in everything, but still...). Multi GPU sure, but not single. Some of the claimed gains understandably had people excited. It was supposed to be the Mantle killer. The hype and PR were just over the top, which is too bad because it detracts from what is an excellent performance driver.

I havnt personally tried those drivers yet. But if we look at pcgameshardware.de results what is it that qualifies this as a different driver than prior nv drivers?

Tell me because i just cant see it? There is ups. But there is certainly also downs. Where is the unusual?

imo Its progress as usual that is appriciated. And its free. Thank you. But it is not like nv drivers have been standing still the last half a year. We have had several solid improvement like this earlier. And we always have had.

Even for nv this is way more bs than usual. Spin and pr.

 

[PinchedNerve]

My PC

335.23

337.50

---------------------------------------------------------------------------------------------------------------------------------------


My Kids PC's.

PC-1. Win 8.1 Pro, i5-3570k, 16 gigs Mem, MSI GTX 770 Lightning.

FarCry 2 - Dx10.
Settings: Demo(Ranch Small), 1920x1080 (60Hz), D3D10, Fixed Time Step(No), Disable Artificial Intelligence(No), Full Screen, Anti-Aliasing(8x), VSync(No), Overall Quality(Ultra High), Vegetation(Very High), Shading(Ultra High), Terrain(Ultra High), Geometry(Ultra High), Post FX(High), Texture(Ultra High), Shadow(Ultra High), Ambient(High), Hdr(Yes), Bloom(Yes), Fire(Very High), Physics(Very High), RealTrees(Very High)
335.23
Average Framerate: 140.37
Max. Framerate: 221.05
Min. Framerate: 106.90

337.50
Average Framerate: 144.85
Max. Framerate: 225.76
Min. Framerate: 109.53

Heaven
Platform: Windows NT 6.2 (build 9200) 64bit
CPU model: Intel(R) Core(TM) i5-3570K CPU @ 3.40GHz (3467MHz) x4
GPU model: NVIDIA GeForce GTX 770 9.18.13.3523 (2048MB) x1
Settings Render: Direct3D11
Mode: 1920x1080 8xAA fullscreen
Preset Custom
Quality Ultra
Tessellation: Extreme
335.23
FPS: 40.3
Score: 1015
Min FPS: 20.0
Max FPS: 91.6

337.50
FPS: 40.1
Score: 1011
Min FPS: 20.0
Max FPS: 89.7

Metro 2033.
Options: Resolution: 1920 x 1080; DirectX: DirectX 11; Quality: High; Antialiasing: MSAA 4X; Texture filtering: AF 16X; Advanced PhysX: Enabled; Tesselation: Enabled; DOF: Enabled
335.23
Average Framerate: 40.00
Max. Framerate: 99.46
Min. Framerate: 9.05

337.50
Average Framerate: 50.00
Max. Framerate: 145.90
Min. Framerate: 9.36

---------------------------------------------------------------------------

PC-2. Win 8.1 Pro, FX 8350, 16 gigs Mem, MSI GTX 770 Lightning.

FarCry 2 - Dx10
Settings: Demo(Ranch Small), 1920x1080 (60Hz), D3D10, Fixed Time Step(No), Disable Artificial Intelligence(No), Full Screen, Anti-Aliasing(8x), VSync(No), Overall Quality(Ultra High), Vegetation(Very High), Shading(Ultra High), Terrain(Ultra High), Geometry(Ultra High), Post FX(High), Texture(Ultra High), Shadow(Ultra High), Ambient(High), Hdr(Yes), Bloom(Yes), Fire(Very High), Physics(Very High), RealTrees(Very High)
335.23
Average Framerate: 106.31
Max. Framerate: 159.42
Min. Framerate: 73.55

337.50
Average Framerate: 111.13
Max. Framerate: 164.47
Min. Framerate: 77.88

Heaven.
Platform: Windows NT 6.2 (build 9200) 64bit
CPU model: AMD FX(tm)-8350 Eight-Core Processor (4334MHz) x4
GPU model: NVIDIA GeForce GTX 770 9.18.13.3523 (2048MB) x1
Settings Render: Direct3D11
Mode: 1920x1080 8xAA fullscreen
Preset Custom
Quality Ultra
Tessellation: Extreme
335.23
FPS: 39.4
Score: 991
Min FPS: 16.5
Max FPS: 89.5

337.50
FPS: 39.0
Score: 983
Min FPS: 15.2
Max FPS: 88.6

Metro 2033.
Options: Resolution: 1920 x 1080; DirectX: DirectX 11; Quality: High; Antialiasing: MSAA 4X; Texture filtering: AF 16X; Advanced PhysX: Enabled; Tesselation: Enabled; DOF: Enabled
335.23
Average Framerate: 37.67
Max. Framerate: 226.14
Min. Framerate: 10.26

337.50
Average Framerate: 38.33
Max. Framerate: 194.89
Min. Framerate: 9.46

Sleeping Dogs.
1920x1080, Settings High, High Res Textures Yes.
335.23
Average Framerate: 90.7
Max. Framerate: 137.8
Min. Framerate: 56.2

337.50
Average Framerate: 96.9
Max. Framerate: 141.0
Min. Framerate: 57.6

 

[rtsurfer]

Pretty much a regular driver release.
Nothing ground breaking to see here.

Maybe now the Nvidia fanboys on this forum will stop hyping this up to be the 'Mantle Killer.'

Cough: cough: blackened23 cough:

Infraction issued for member callout.
-- stahlhart

 

[sxr7171]

My machine froze in a bad way with this driver. Can't be sure it's the driver but I needed to restore to boot back up. The restore put the 335 driver back on. This is Win 8.1 x64.

In my limited testing Hitman Absolution performed better as did Sleeping Dogs. the freeze occurred while testing AC4.

Can anything stop the tearing in AC4? Tried Vsync in game, in driver, and enabled triple buffer in NVinspector. Nothing stops that tearing. Nothing.

 

[PPB]

Now we see why they were so enthusiastic with DX12. Time to jump from the Wonder Driver hype train back to the DX12 one again guys.

 

[PinchedNerve]

For the AMD fanboys poo-pooing the thread. http://www.geforce.com/whats-new/articles/nvidia-geforce-337-50-beta-performance-driver

 

[sxr7171]

For the AMD fanboys poo-pooing the thread. http://www.geforce.com/whats-new/articles/nvidia-geforce-337-50-beta-performance-driver

Some good info there. Nice to see 1080p max settings as that's what I use. Still need to test some of these games myself. Seems like a step in the right direction.


Even though they didn't mention the Frostbite 3 games I'm seeing some pretty big performance increases. BF4 at 145% resolution scaling is running at a solid 60fps. I will try 150% next. NFS rivals used to require 1.5 to run 60fps solid I tried 1.7 and it's running a smooth 60fps. That was not possible before. I'm going to keep increasing those to find the limit.

I have no idea what these optimizations are doing but my guess is that I was VRAM limited at those scaling levels. And this driver may be better at conserving VRAM. I realize most people don't run resolution scaling but I like the crispness it provides - a lot. The goal is to run both games at 200% or 2.0 scaling. Maybe it will take a next gen card to do that. Hopefully not.

 

[Hitman928]

For the AMD fanboys poo-pooing the thread. http://www.geforce.com/whats-new/articles/nvidia-geforce-337-50-beta-performance-driver

Honest question, what's different in that link than what nvidia had already released in promoting the driver? I quickly went through it and didn't really see anything new except maybe a couple new games included in their graphs?

 

[sxr7171]

Honest question, what's different in that link than what nvidia had already released in promoting the driver? I quickly went through it and didn't really see anything new except maybe a couple new games included in their graphs?

Actual game settings as opposed to maximum increase of X% with unknown settings.

 

[krumme]

For the AMD fanboys poo-pooing the thread. http://www.geforce.com/whats-new/articles/nvidia-geforce-337-50-beta-performance-driver

Lots of technical explanation like:

"By remaining laser-focused on DirectX 11 (and the forthcoming DirectX 12), we have dramatically enhanced the efficiency of our DirectX 11 driver"

So they are using laser tech to remain focused on dx11 (and dx12 at the same time)

Seriously. If people said to them to shut up and spare the pr they could use more time on the drivers instead of making up this nonsense.

 

[krumme]

My PC

335.23

337.50

---------------------------------------------------------------------------------------------------------------------------------------


My Kids PC's.

PC-1. Win 8.1 Pro, i5-3570k, 16 gigs Mem, MSI GTX 770 Lightning.

FarCry 2 - Dx10.
Settings: Demo(Ranch Small), 1920x1080 (60Hz), D3D10, Fixed Time Step(No), Disable Artificial Intelligence(No), Full Screen, Anti-Aliasing(8x), VSync(No), Overall Quality(Ultra High), Vegetation(Very High), Shading(Ultra High), Terrain(Ultra High), Geometry(Ultra High), Post FX(High), Texture(Ultra High), Shadow(Ultra High), Ambient(High), Hdr(Yes), Bloom(Yes), Fire(Very High), Physics(Very High), RealTrees(Very High)
335.23
Average Framerate: 140.37
Max. Framerate: 221.05
Min. Framerate: 106.90

337.50
Average Framerate: 144.85
Max. Framerate: 225.76
Min. Framerate: 109.53

Heaven
Platform: Windows NT 6.2 (build 9200) 64bit
CPU model: Intel(R) Core(TM) i5-3570K CPU @ 3.40GHz (3467MHz) x4
GPU model: NVIDIA GeForce GTX 770 9.18.13.3523 (2048MB) x1
Settings Render: Direct3D11
Mode: 1920x1080 8xAA fullscreen
Preset Custom
Quality Ultra
Tessellation: Extreme
335.23
FPS: 40.3
Score: 1015
Min FPS: 20.0
Max FPS: 91.6

337.50
FPS: 40.1
Score: 1011
Min FPS: 20.0
Max FPS: 89.7

Metro 2033.
Options: Resolution: 1920 x 1080; DirectX: DirectX 11; Quality: High; Antialiasing: MSAA 4X; Texture filtering: AF 16X; Advanced PhysX: Enabled; Tesselation: Enabled; DOF: Enabled
335.23
Average Framerate: 40.00
Max. Framerate: 99.46
Min. Framerate: 9.05

337.50
Average Framerate: 50.00
Max. Framerate: 145.90
Min. Framerate: 9.36

---------------------------------------------------------------------------

PC-2. Win 8.1 Pro, FX 8350, 16 gigs Mem, MSI GTX 770 Lightning.

FarCry 2 - Dx10
Settings: Demo(Ranch Small), 1920x1080 (60Hz), D3D10, Fixed Time Step(No), Disable Artificial Intelligence(No), Full Screen, Anti-Aliasing(8x), VSync(No), Overall Quality(Ultra High), Vegetation(Very High), Shading(Ultra High), Terrain(Ultra High), Geometry(Ultra High), Post FX(High), Texture(Ultra High), Shadow(Ultra High), Ambient(High), Hdr(Yes), Bloom(Yes), Fire(Very High), Physics(Very High), RealTrees(Very High)
335.23
Average Framerate: 106.31
Max. Framerate: 159.42
Min. Framerate: 73.55

337.50
Average Framerate: 111.13
Max. Framerate: 164.47
Min. Framerate: 77.88

Heaven.
Platform: Windows NT 6.2 (build 9200) 64bit
CPU model: AMD FX(tm)-8350 Eight-Core Processor (4334MHz) x4
GPU model: NVIDIA GeForce GTX 770 9.18.13.3523 (2048MB) x1
Settings Render: Direct3D11
Mode: 1920x1080 8xAA fullscreen
Preset Custom
Quality Ultra
Tessellation: Extreme
335.23
FPS: 39.4
Score: 991
Min FPS: 16.5
Max FPS: 89.5

337.50
FPS: 39.0
Score: 983
Min FPS: 15.2
Max FPS: 88.6

Metro 2033.
Options: Resolution: 1920 x 1080; DirectX: DirectX 11; Quality: High; Antialiasing: MSAA 4X; Texture filtering: AF 16X; Advanced PhysX: Enabled; Tesselation: Enabled; DOF: Enabled
335.23
Average Framerate: 37.67
Max. Framerate: 226.14
Min. Framerate: 10.26

337.50
Average Framerate: 38.33
Max. Framerate: 194.89
Min. Framerate: 9.46

Sleeping Dogs.
1920x1080, Settings High, High Res Textures Yes.
335.23
Average Framerate: 90.7
Max. Framerate: 137.8
Min. Framerate: 56.2

337.50
Average Framerate: 96.9
Max. Framerate: 141.0
Min. Framerate: 57.6

How am i to understand those metro results for the kids pc intel vs amd cpu. Damn !

 

[Keysplayr]

Lots of technical explanation like:

"By remaining laser-focused on DirectX 11 (and the forthcoming DirectX 12), we have dramatically enhanced the efficiency of our DirectX 11 driver"

So they are using laser tech to remain focused on dx11 (and dx12 at the same time)

Seriously. If people said to them to shut up and spare the pr they could use more time on the drivers instead of making up this nonsense.

I think by saying they're laser focused, they mean very sharply focused. Not actually using laser technology on DX technology. ()

 

[PinchedNerve]

I think by saying they're laser focused, they mean very sharply focused. Not actually using laser technology on DX technology. ()

I would hope English isn't his first language when Interpreting "Laser Focus" as "Laser Technology".

 

[sxr7171]

Okay so I tested NFS Rivals and BF4.

BF4 now I can do 150% scaling at solid 60fps.
NFS Rivals 175% scaling at solid 60fps.

It may be in my head but scaling to nice 125/150/175/200% reduces visible jaggies compared to 145/160/165% etc.

I tried 200% scaling in NFS Rivals and that was a no-go. It looked amazing though.

 

[SirPauly]

Lots of technical explanation like:

"By remaining laser-focused on DirectX 11 (and the forthcoming DirectX 12), we have dramatically enhanced the efficiency of our DirectX 11 driver"

So they are using laser tech to remain focused on dx11 (and dx12 at the same time)

Seriously. If people said to them to shut up and spare the pr they could use more time on the drivers instead of making up this nonsense.

You may be misreading and making things up!

nVidia may be very focused on DirectX efficiency!

 

[SirPauly]

Thought this was interesting:

As with 337.50 Beta’s single-GPU DirectX 11 optimizations, CPU overhead reductions apply globally, benefitting the majority of DirectX 9, DirectX 10, and DirectX 11 games

http://www.geforce.com/whats-new/articles/nvidia-geforce-337-50-beta-performance-driver

 

[sxr7171]

Crysis 3 maxed out 1080p TXAA 4x 60fps with some dips to mid/high 50s. Very playable maxed out finally.

Metro Last Light solid 60fps with 2x SSAA.

780ti OC +130MHz. VRAM +400MHz.

Btw Crysis 3 looks amazing.

Edit: Unfortunately I got "video scheduler internal error" might be the OC, might be these drivers need a little work. I've never seen that error before.

 

[Mondozei]

Even as someone who has an AMD GPU I welcome these news. If Nvidia is raising the bar, then is good for everyone(even if AMD has in some sense raised the bar already with Mantle, its still only a select few games).

But the larger point here is unmistakable, would Nvidia be as obsessed with drivers if they hadn't the new Maxwell GPUs coming out? They've always had a (perceived) driver advantage, and not without reason, but this new energy they have, I think, is perhaps less due to Mantle and more due to the fact that they basically will have little to nothing to offer this year for potential customers than same old, same old, just like AMD.

And to encourage people to buy old tech you have to do what the console industry has done much better than the PC industry: optimize, optimize and optimize.
(Even putting aside the vast diversity of hardware, the PC companies should have been able to do this job much better than it did historically, which DX 12 development is a inadvertantly admitting).
Now, the counter-argument would be that they can walk and chew at the same time. Maybe, but that hasn't historically been the record. The PC divisions of these companies have typically been very bad at driver optimizations, even if Nvidia has typically been better than AMD it hasn't been awesome.

That'll hopefully change with DX 12 and/or as Mantle rolls out further and further, that these optimizations will happen much more rapidly and naturally, but there's still some time to go for that.

All in all, good news, but the backdrop here is a stagnating GPU market due to endless delays on the new nodes.

 

[Despoiler]

I'm not seeing anything "approaching Mantle like performance". If you just look at max and average FPS in the Nvidia marketing slides or review sites, then sure. However, if you look at the min FPS numbers you see sub 5% gains or even negative gains depending on the game. Mantle slashes frame times by half and as such boosts min FPS by 20-100% depending on the system and game. That is the true measure of Mantle's efficiency. The Great Frame Time Debate, as it came to be known, raised the bar for measuring performance. Yet here we are again with Nvidia touting max and average FPS.

 

[VulgarDisplay]

This does make me feel some remorse for returning my gtx780 for an r9 290, but bf4 mantle performance was my deciding factor.

Much like the mantle PR this shows that you really need to temper your expectations when that magical "up to" statement appears. Some nice gains, but not nearly as groundbreaking as we were led to believe.

 

[PinchedNerve]

We’re not quite finished, however, as many more changes, optimizations, and enhancements are waiting in the wings for integration into our next WHQL release.

http://www.geforce.com/whats-new/articles/nvidia-geforce-337-50-beta-performance-driver

 

[Gloomy]

http://www.geforce.com/whats-new/articles/nvidia-geforce-337-50-beta-performance-driver

empty PR

 

[krumme]

I think by saying they're laser focused, they mean very sharply focused. Not actually using laser technology on DX technology. ()

Well ofcource not. - its not an eye laser operation. Did you really miss my point?

Its just pr nonsense. Writing you are focused on one thing (dx11) and in the same sentence writing you are focused on another (dx12). When you do that you actually write you are focusing on two different things. Dx11 and dx12/mantle is as we know very different. How is that sharply focused?

When amd introduced the 14 beta driver with framepacing, mantle, dx and ggpu capabilities were they then working titanium, uber, laser focused? No it was excactly oposite. They were working on different technologies. As nv is with both higher and lower level api. So when they say laser focused its just typical nv pr because they are actually doing the exact oposite, - for good reasons.

 

[Attic]

This does make me feel some remorse for returning my gtx780 for an r9 290, but bf4 mantle performance was my deciding factor.

Much like the mantle PR this shows that you really need to temper your expectations when that magical "up to" statement appears. Some nice gains, but not nearly as groundbreaking as we were led to believe.

Yep.

It ought to be troubling when we have to assume that everything we hear from any NV PR or AMD PR is being said for effect, and not for the straightforward expression of an honest opinion or expected and real results.

The word we're really looking for with the PR spin from both companies is, Sham.

I'd call what nVidia did here with their early release of performance numbers and how they went about revealing this release driver to be a sham. Hard to conclude anything other than the sham works given it has continued to the degree it does.

 

[guskline]

Pinchednerve, what are the specs of YOUR machine. I saw the one of your child. I know you have a GTX780 but what cpu etc and is your gpu OC'd?