Nvidia Releases GeForce 314.22 WHQL Drivers

Toggle sidebar Toggle sidebar
Status
Not open for further replies.
Pole

Pole

Honorable
Mar 24, 2013
42
0
10,530
60% faster on 680 for tomb raider, thats amazing !

So GTX 670 falls far behind on tomb raider ?
 
razor512

razor512

Distinguished
Jun 16, 2007
2,134
71
19,890
Please tomshardware, do a review where you test nvidia and even ATI driver improvements across GPU generations.

For example with an improvement on a GTX 680 translate to a improvement on a GTX 580, 480, 280, and probably even a 9800.

While everyone is going to update anyway, it would be good to know if users on older hardware are gaining anything at all, and if so, then how much of an improvement is to be had.
 
F

FLanighan

Distinguished
Jul 19, 2009
180
0
18,690
[citation][nom]thedude300[/nom]I feel like the 660 and 670 have just been forgotten in these updates.[/citation]

I have a 670 and noticed a huge improvement in tomb raider.
 
griptwister

griptwister

Distinguished
Oct 7, 2012
1,437
0
19,460
In case you guys didn't notice... These drivers are for the 6xx series cards... they just did the testing with a 680. They listed they used a 680 so there won't be that one guy who says, "I'm Using a GT 630 and I only get one FPS more!!! Nvidia Lied to us!" Lol!
 
dragonsqrrl

dragonsqrrl

Distinguished
Nov 19, 2009
1,280
0
19,290
[citation][nom]thedude300[/nom]I feel like the 660 and 670 have just been forgotten in these updates.[/citation]
"Increases performance for GeForce 400/500/600 series GPUs in several PC games... Here are a few examples:"
 
fkr

fkr

Splendid
May 24, 2009
2,646
0
21,460
yep these show improvements across the whole series of cards. I love how this comes out the day before bioshock is released. thanks for the support nvidia.
 
omnimodis78

omnimodis78

Distinguished
Oct 7, 2008
886
0
19,010
To be honest, I'm also a little annoyed by the 680 benchmark nvidia keeps quoting in their improvements. If the drivers are tuned, shouldn't the performance gain be essentially the same percentage all across the board? So a 40% performance increase on a 680 doesn't mean a 40% improvement on a 670, or a 660? Maybe someone who actually has the knowledge can articulate why this point is wrong.
 
razor512

razor512

Distinguished
Jun 16, 2007
2,134
71
19,890
[citation][nom]dragonsqrrl[/nom]"Increases performance for GeForce 400/500/600 series GPU's in several PC games... Here are a few examples:"[/citation]

only problem is they always list that but then they also listed that it was tested on a 680 (they are not actually testing the older cards to see if the optimizations apply to the older cards.
 
B

blazorthon

Glorious
Sep 24, 2010
14,022
7
43,265
[citation][nom]omnimodis78[/nom]To be honest, I'm also a little annoyed by the 680 benchmark nvidia keeps quoting in their improvements. If the drivers are tuned, shouldn't the performance gain be essentially the same percentage all across the board? So a 40% performance increase on a 680 doesn't mean a 40% improvement on a 670, or a 660? Maybe someone who actually has the knowledge can articulate why this point is wrong.[/citation]

Unfortunately, that is wrong and the true answer is a lot more complex. The higher end a card is, the more it is affected by driver improvements. Chances are that if we go down to something that's about half as fast as the GTX 680, say either the GTX 650 Ti or the GTX 650 Ti Boos,t the difference will probably be about half, maybe less. Different ratios of memory performance to GPU performance as well as GPU core count to GPU frequency can also affect how much diver improvements scale between cards.

Whether or not the CPU and the rest of the system is fast enough to take advantage of the gains can also impact the improvements.
 
B

blazorthon

Glorious
Sep 24, 2010
14,022
7
43,265
According to Tom's this driver introduces bad stutter issues for X79 systems as shown and explained in Tom's review of the GTX 650 Ti Boost. Could we get some more information on this? Nvidia says that they can't reproduce the issue, but IDK if they tested in a similar system or not. Is the issue just Tom's system starting to fail or some other fault with it, or is it more wide-spread?

EDIT: My bad, that was 314.21. Oops. Well then, it's worth asking Tom's this: Does this version fix the stutter issue?
 
17seconds

17seconds

Champion
Jul 27, 2010
10,584
4
51,965
[citation][nom]omnimodis78[/nom]To be honest, I'm also a little annoyed by the 680 benchmark nvidia keeps quoting in their improvements. If the drivers are tuned, shouldn't the performance gain be essentially the same percentage all across the board? So a 40% performance increase on a 680 doesn't mean a 40% improvement on a 670, or a 660? Maybe someone who actually has the knowledge can articulate why this point is wrong.[/citation]

It seems like you and others have taken the release notes too literally. Please do read the notes carefully:

"Increases performance for Geforce 400/500/600 series" - Means the GTX 670 and 660 will also see performance gains. In fact, means just about any card sold over the past 5 years could see performance gains.

"Results will vary depending on your GPU and system configuration" - Means that not everyone will see the same level of improvement. The amount of the performance gain could be the same, could be less, or it could even be more.

"Up to 41% improvement in Bioshock Infinite" - Means anything between 1% - 41%. It does not literally mean that every single GTX 680 user will only and always experience exactly a 41% boost in that game. It also says nothing about what to expect with other GPU's. Again, the gains could be more, could be less, or could be exactly the same.
 
B

bystander

Champion
Dec 9, 2009
15,141
6
57,965
[citation][nom]matto17secs[/nom]It seems like you and others have taken the release notes too literally. Please do read the notes carefully:"Increases performance for Geforce 400/500/600 series" - Means the GTX 670 and 660 will also see performance gains. In fact, means just about any card sold over the past 5 years could see performance gains."Results will vary depending on your GPU and system configuration" - Means that not everyone will see the same level of improvement. The amount of the performance gain could be the same, could be less, or it could even be more."Up to 41% improvement in Bioshock Infinite" - Means anything between 1% - 41%. It does not literally mean that every single GTX 680 user will only and always experience exactly a 41% boost in that game. It also says nothing about what to expect with other GPU's. Again, the gains could be more, could be less, or could be exactly the same.[/citation]
I don't think you can expect more. They did say "up to", not "about", but yeah, you can't expect to get those exact numbers.
 
dragonsqrrl

dragonsqrrl

Distinguished
Nov 19, 2009
1,280
0
19,290
[citation][nom]razor512[/nom]only problem is they always list that but then they also listed that it was tested on a 680 (they are not actually testing the older cards to see if the optimizations apply to the older cards.[/citation]
If what you're saying were the case, I would expect to see the performance of older hardware (1 or 2 generations back) drop off a cliff in new AAA games relative to current gen hardware. But I'm not sure I've ever seen this happen. It's standard fair to optimize for the latest generation, and to devote the most resources into driver development for the latest and greatest, but just because performance highlights are listed only for the 680 doesn't necessarily mean they aren't devoting some resources to older products. It's not uncommon to see driver optimizations such as these positively effecting previous generation hardware, especially considering the architectural similarities between Fermi and Kepler. So I would argue that the opposite of what you're suggesting is more likely the case. With a unified driver model Nvidia has an obligation to maintain functionality on all the products their latest drivers claim to support.
 
spartanmk2

spartanmk2

Honorable
May 11, 2012
470
0
10,790
Only found out about the new driver after HotS crashed during a cutscene and informed me my driver may be out of date.

After installing, noticed a pretty good jump in fps (10-15).
 
ericjohn004

ericjohn004

Honorable
Oct 26, 2012
651
0
11,010
You gotta love Nvidia for their driver updates. I mean seriously, they JUST came out with one a couple weeks ago, and again a month or so before that. I've only had my 660 Ti since late December and I've already has to update my driver 3 times. Once the first time, then the second, and now today.

If only AMD card owners were so lucky.
 
17seconds

17seconds

Champion
Jul 27, 2010
10,584
4
51,965

I guess you stick to the official WHQL drivers. If you include the beta drivers, I count 7 official driver releases since December for my GTX 580.
 
shadowfamicom

shadowfamicom

Honorable
May 2, 2012
136
0
10,680
[citation][nom]razor512[/nom]Please tomshardware, do a review where you test nvidia and even ATI driver improvements across GPU generations.For example with an improvement on a GTX 680 translate to a improvement on a GTX 580, 480, 280, and probably even a 9800.While everyone is going to update anyway, it would be good to know if users on older hardware are gaining anything at all, and if so, then how much of an improvement is to be had.[/citation]

Yeah, always wondering if these updates will do anything for one of my rigs that still has a 295 GTX in it, and another that has a 9800 GTX+
 
17seconds

17seconds

Champion
Jul 27, 2010
10,584
4
51,965

I would assume so, since many of the enhancements are game specific. There's only one way to find out....
 
Pole

Pole

Honorable
Mar 24, 2013
42
0
10,530
After this update
Sapphire Radeon HD 7970 GHz Edition 3GB vs GTX 670 4GB
Which one would you go for?
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom