NVidia driver 337.50 beta released!

[Deders]

nm

 

[96Firebird]

Lies.

Did he say in SLI? :whiste:

 

[Face2Face]

Ran some benchmarks last night to see if these new 337.50 drivers would give me a frame-rate boost in any of these games below.

To my surprise.....they.... well....

These numbers are with the rig in the sig, and the Lightning overclocked to 1306/1656


Tomb Raider

1440p - Ultimate settings

335.23 - Avg FPS - 64.2
337.50 - Avg FPS - 64.4

Metro Last Light

1440p - Very high settings - DX11 -PhysX - NO SSAA

335.23 - Avg FPS - 48.08
337.50 - Avg FPS - 49.41

Metro 2033

1440p - Very high settings - 4XMSAA - PhysX

335.23 - Avg FPS - 33.66
337.50 - Avg FPS - 32.13

Bioshock Infinite

1440p - DX11 Ultra- DOF

335.23 - Avg FPS - 81.1
337.50 - Avg FPS - 82.57

Hitman Absolution

1440p - Ultra settings - 2XMSAA

335.23 - Avg FPS - 60.82
337.50 - Avg FPS - 61.45

Batman Arkham Orgins

1440p - settings maxed - TXAA High - PhysX

335.23 - Avg FPS - 58
337.50 - Avg FPS - 59

GTA IV

Modded with High resolution textures and ENB

1440p - Maxed out and sliders all the way up

335.23 - Avg FPS - 58.76
337.50 - Avg FPS - 66.05


Decent gains for GTA IV, but everything else is pretty meh... :|

 

[railven]

Did he say in SLI? :whiste:

Did he say single card? :whiste:

 

[railven]

Ran some benchmarks last night to see if these new 337.50 drivers would give me a frame-rate boost in any of these games below.

To my surprise.....they.... well....

These numbers are with the rig in the sig, and the Lightning overclocked to 1306/1656


Tomb Raider

1440p - Ultimate settings

335.23 - Avg FPS - 64.2
337.50 - Avg FPS - 64.4

Metro Last Light

1440p - Very high settings - DX11 -PhysX - NO SSAA

335.23 - Avg FPS - 48.08
337.50 - Avg FPS - 49.41

Metro 2033

1440p - Very high settings - 4XMSAA - PhysX

335.23 - Avg FPS - 33.66
337.50 - Avg FPS - 32.13

Bioshock Infinite

1440p - DX11 Ultra- DOF

335.23 - Avg FPS - 81.1
337.50 - Avg FPS - 82.57

Hitman Absolution

1440p - Ultra settings - 2XMSAA

335.23 - Avg FPS - 60.82
337.50 - Avg FPS - 61.45

Batman Arkham Orgins

1440p - settings maxed - TXAA High - PhysX

335.23 - Avg FPS - 58
337.50 - Avg FPS - 59

GTA IV

Modded with High resolution textures and ENB

1440p - Maxed out and sliders all the way up

335.23 - Avg FPS - 58.76
337.50 - Avg FPS - 66.05


Decent gains for GTA IV, but everything else is pretty meh... :|

With what Pauly said, it seems these drivers focus on CPU bottlenecks. My guess is all your examples above you are GPU limited?

Oh well, I'm back to one GTX 660 Ti for the time being, I'm not going to bother testing haha.

 

[Grooveriding]

He did say he was kidding.

To be fair though, Groove's been called worse.

Edit: Since He's bought 2x Titans + 2x 780's + 2x 780 ti's this gen alone, I'd say he has a right to stating his opinion and not have it be considered bias. He's disappointed.

Wasn't even disappointment. I am happy with my cards. Just calling out the obvious BS and PR that was pushed around about these drivers and them being called 'super drivers' from nvidia and a lot of the claimed performance increases being nonsense. We just have the same issue here on this forum of a small fringe where the purchase of a video card must be tantamount to total devotion as mandatory with the purchase. Not a big deal, we've seen much worse here come of that same broken down thought process.

Also we now have reviewers seeing that these drivers raised the temperature throttle, allowing cards to run at a higher boost clock, allowing for more performance... A lot of hype and PR is mostly what these drivers were, along with a stealth performance bump in the form of a temperature throttle increase.

Lies.

<snip>

Windows 8.1, latest drivers. i7 3770K @ 4.8ghz (extreme OC to reduce CPU bottleneck too).
http://hardocp.com/article/2014/04/08/amd_radeon_r9_295x2_video_card_review/2

Multiplayer is where Mantle dominates.

I was under the impression that crossfire still scales better than SLI does ? I don't know. You see a lot of FUD tossed around about drivers and DX11 MT in BF4. Mostly based around a set of broken benchmarks run at that pclab website on the old beta client of the game.

Always using the 780ti vs 290X example. Somehow many forget that the 780ti is simply a faster card than the 290X. It also overclocks better for the most part and sees more performance gains from overclocking than the 780ti. So it makes sense that it shows as faster in BF4, because gasp, it's a faster card.

I would assume 295X2 vs 780ti is showing better numbers for 295X2 because of Mantle and Crossfire showing better scaling than SLI does. SLI sucks in BF4 anyways, it's one of the few games where it does not feel smooth and at times a single card feels a lot better. It helps the minimums a lot, but it feels like nvidia's frame pacing is way out of whack in Battlefield 4.

 

[Silverforce11]

Did he say in SLI? :whiste:

He simply said it was faster, which is false based on evidence from a reputable site.

Unless you can show some evidence of the latest drivers in action from other reputable sites in BF4 MP, his statement is a lie.

 

[Silverforce11]

Also we now have reviewers seeing that these drivers raised the temperature throttle, allowing cards to run at a higher boost clock, allowing for more performance... A lot of hype and PR is mostly what these drivers were, along with a stealth performance bump in the form of a temperature throttle increase.

I was under the impression that crossfire still scales better than SLI does ? I don't know. You see a lot of FUD tossed around about drivers and DX11 MT in BF4. Mostly based around a set of broken benchmarks run at that pclab website on the old beta client of the game.

Always using the 780ti vs 290X example. Somehow many forget that the 780ti is simply a faster card than the 290X. It also overclocks better for the most part and sees more performance gains from overclocking than the 780ti. So it makes sense that it shows as faster in BF4, because gasp, it's a faster card.

I would assume 295X2 vs 780ti is showing better numbers for 295X2 because of Mantle and Crossfire showing better scaling than SLI does. SLI sucks in BF4 anyways, it's one of the few games where it does not feel smooth and at times a single card feels a lot better. It helps the minimums a lot, but it feels like nvidia's frame pacing is way out of whack in Battlefield 4.

It's not just BF4 with Mantle.

It's the same in Tomb Raider, Far Cry 3 and even Crysis 3: 20% faster in these games at 4K.

Terrible dips in fps for the 780ti SLI:

http://hardocp.com/article/2014/04/08/amd_radeon_r9_295x2_video_card_review/5#.U0VW7fmSxSk

As for which card is faster now overall, these latest drivers from both NV and AMD doing battle in a lot of games:
http://www.computerbase.de/2014-04/amd-radeon-r9-295x2-benchmark-test/5/

Note that they tested Mantle and had a 48% performance boost in BF4 MP, but from what I can understand, it was an extra data not included in their overall chart since they compared the DX11 results. Looks like single card with both latest drivers, R290X caught up with the 780ti. SLI from this result scales fine, though a few games where CF was broken skew the outcome.

Interestingly, when the 780ti is set for max boost with increased power target and 100% fanspeed:

Edit: Yup, Computerbase.de took BF4 SP results and DX11 for their summary chart, ignoring the massive advantage of Mantle in BF4 MP: http://www.computerbase.de/2014-04/amd-radeon-r9-295x2-benchmark-test/14/ Including the MP result would skew the chart in favor of R290X vs 780ti and likewise for CF vs SLI.

 

[krumme]

He simply said it was faster, which is false based on evidence from a reputable site.

Unless you can show some evidence of the latest drivers in action from other reputable sites in BF4 MP, his statement is a lie.

I dont doubt a single 780ti is in general a faster card but if we look at pcper latest test in bf4 the 290x shows less frame variance on the last percentile than the 780ti and thats even accounting for the 290x beeing 1050ghz. Thats for 1440 and 4k resolutions. And in single player mode.

Now if you dont want min fps below 60fps in MP mantle is simple a requirement unless you have a 6 core or run your cpu at 4.8Ghz. So its a personal preference deciding what your requirement are vs your cpu power. It is subjective. If you want very high fps you are simply cpu limited at one point.

Personally for bf4 there is no way i am playing on my ib 4ghz without mantle because i am adicted to the high fps and playing bad without. I will always skimp on the gfx details to get there.

But my kid is different there. He doesnt give a damn about either gfx details and very high fps. He always complain about the netcode. Lol. You can imagine who is the best player.

 

[96Firebird]

He simply said it was faster, which is false based on evidence from a reputable site.

Unless you can show some evidence of the latest drivers in action from other reputable sites in BF4 MP, his statement is a lie.

He did, and it is faster in single-card. So he isn't lying...

Plus, he said all that before the reviews for the 295x2 were even posted, and that includes numbers from a driver that isn't available yet. So at his time of posting, he wasn't lying.

 

[krumme]

He did, and it is faster in single-card. So he isn't lying...

Plus, he said all that before the reviews for the 295x2 were even posted, and that includes numbers from a driver that isn't available yet. So at his time of posting, he wasn't lying.

No its not faster in - bf4 - as he said. Its slower. So its wrong.

But people can please themselves. If they want to play sp bf4 its a free choise but i doubt the relevance as mp is where the bottlenecks is.

 

[sontin]

He did, and it is faster in single-card. So he isn't lying...

Plus, he said all that before the reviews for the 295x2 were even posted, and that includes numbers from a driver that isn't available yet. So at his time of posting, he wasn't lying.

He compares a 600W "single" card with 2x 200W single cards and claims that i lied because the 600W monster is still faster with Mantle. :|

 

[Gloomy]

He compares a 600W "single" card with 2x 200W single cards and claims that i lied because the 600W monster is still faster with Mantle. :|

780Ti is a 250 watt card per Nvidia's specifications. I'm sure you knew that though.

 

[Majcric]

Anybody have any driver comparison results for Company of Heroes 2?

 

[ams23]

Lies.

GTX 780 Ti SLI tends to be a bit faster at 25x16 (and lower resolutions) with high details:

...while also having better frametime delta percentages:

 

[Silverforce11]

GTX 780 Ti SLI tends to be a bit faster at 25x16 (and lower resolutions) with high details:

...while also having better frametime delta percentages:

In Singleplayer. Sure.

As for the lame power use argument, even cranked up, SLI 780ti is still slower than CF R290X or 295 X2 in BF4 Multiplayer while using similar power.

 

[ams23]

Well you aren't making much sense. You are talking about BF4 performance, and then you list a Crysis 3 power consumption chart.

GTX 780 Ti SLI absolutely dominates 295X2 in Crysis 3 perf. per watt in Anandtech's test suite:

 

[Silverforce11]

Nope, its just a measure from Computerbase.de setting 100% fanspeed and max power target for the 780ti when they bench it. The point is the power use jumps from 471W to 628W. I just think its funny they do this and do not include the BF4 MP results with Mantle in their summary charts but only their SP benches.

 

[ams23]

From the perspective of a hardware reviewer that has to review one video card after another after another, single player benchmarks make sense because they tend to be more repeatable from one test run to another, and network speed variability doesn't come into play.

 

[3DVagabond]

Wasn't even disappointment. I am happy with my cards. Just calling out the obvious BS and PR that was pushed around about these drivers and them being called 'super drivers' from nvidia and a lot of the claimed performance increases being nonsense. We just have the same issue here on this forum of a small fringe where the purchase of a video card must be tantamount to total devotion as mandatory with the purchase. Not a big deal, we've seen much worse here come of that same broken down thought process.

Sorry, I wasn't trying to say you were disappointed with your cards. I assumed that after the way these drivers were hyped that the actual performance would have been a let down.

 

[krumme]

From the perspective of a hardware reviewer that has to review one video card after another after another, single player benchmarks make sense because they tend to be more repeatable from one test run to another, and network speed variability doesn't come into play.

Reviews are for gamers. Thats their purpose. Reviews should reflect gamer behavior.

Useless results is useless no matter how precise they are.

 

[guskline]

If I've learned anything by being a reader and poster of this forum it is that reviews, if done properly, are a guide to performance, not an absolute. It's sad that we have to resort to calling someone a "liar" when they express an opinion that differs with what we believe.

I post my specs of my rig (actually have 3 but due to space limits in the sig space I culled it to one). I love these threads when posters actually run their own tests and post their own results with their own specs. Where I struggle is when someone quotes a "source" to advance their preference, then extrapolates their opinion that their card is better, faster etc and doesn't even run the benchmarks on their system!

It would be like me saying my single EVGA GTX780 Classified, watercooled & Overclocked is faster than an AMD R9 295 because I found one test result run on the R9 295 that I just edged out. Who's kidding whom? My GTX780 will always be in the rear view mirror of a R9 295 and well it should.

Honestly I was about to buy a R9 290 and water cool it to replace my 2 GTX670s (now in my 3770k rig) but the price jumped so high so fast that I could "justify" a 780 Classified. I'm thrilled with this card. On benchmarks it almost ties my 3770k rig with 2 670s in sli and feels smoother to boot.

Now that the mining craze has eased some and prices seem to be adjusting the R9 290 appears to be a better buy for the $$.

Monitor resolution plays a big part in this discussion. The results appear to favor AMD as the resolution increases, especially 4k.

As to the topic that the OP posted, the release of these new drivers, my experience with my main rig doesn't show much if any gain and I had problems running 3d Mark so I'll wait until the official release.

 

[omeds]

Note that they tested Mantle and had a 48% performance boost in BF4 MP, but from what I can understand, it was an extra data not included in their overall chart since they compared the DX11 results. Looks like single card with both latest drivers, R290X caught up with the 780ti. SLI from this result scales fine, though a few games where CF was broken skew the outcome.


Edit: Yup, Computerbase.de took BF4 SP results and DX11 for their summary chart, ignoring the massive advantage of Mantle in BF4 MP: http://www.computerbase.de/2014-04/amd-radeon-r9-295x2-benchmark-test/14/ Including the MP result would skew the chart in favor of R290X vs 780ti and likewise for CF vs SLI.

Sure, but they only benched under win7. GTX780ti SLI would see a large gain under win8.1 in BF4, from similar overhead efficiency gains as Mantle. I get ~50% boost on my vanilla 780's.

 

[norseamd]

seems that you guys might have already figured this out but extreme tech had a article on the lack of promised advancements from this driver release

http://www.extremetech.com/gaming/1...-shouldnt-trust-manufacturer-provided-numbers

 

[Silverforce11]

Sure, but they only benched under win7. GTX780ti SLI would see a large gain under win8.1 in BF4, from similar overhead efficiency gains as Mantle. I get ~50% boost on my vanilla 780's.

Win 8.1 with i7 @ 4.8Ghz

Similar numbers to Computerbase.de MP test at 4K.