NVIDIA NEW DRIVER KILLS MANTLE IN BF4!!

[geogga]

I love when companies make graphs because I just laugh at how misleading it is and how some people just shit their pants seeing the huge improvement which was in reality a small marginal improvement.

Oh and love the huge pictures and fanboyism including the hugely misleading title

[gastew15]

Well considering BF4 runs awful in Mantle it's not really too surprising...

[Mitrailleuse]

i did not get any boost on my GTX 550 ti:(

550 ti is fermi architecture,

this update is for kepler architecture.(6XX,7XX)

[ViruZ_HUN]

550 ti is fermi architecture,

this update is for kepler architecture.(6XX,7XX)

so evil...

[KakaoDj]

They have stretched the graphs to make amd products look like shit. These nvidia marketing guys are assholes.

yeah, like every other company ever. Seriously, don't give Nvidia shit for doing that.

[beebskadoo]

550 ti is fermi architecture,

this update is for kepler architecture.(6XX,7XX)

Yeah....about that...

 

Supported Products

 

ION (Desktops)ION ION LE (Desktops)ION LE GeForce 700 SeriesGeForce GTX TITAN Black, GeForce GTX TITAN, GeForce GTX 780 Ti, GeForce GTX 780, GeForce GTX 770, GeForce GTX 760, GeForce GTX 760 Ti (OEM), GeForce GTX 750 Ti, GeForce GTX 750, GeForce GTX 745 GeForce 600 SeriesGeForce GTX 690, GeForce GTX 680, GeForce GTX 670, GeForce GTX 660 Ti, GeForce GTX 660, GeForce GTX 650 Ti BOOST, GeForce GTX 650 Ti, GeForce GTX 650, GeForce GTX 645, GeForce GT 645, GeForce GT 640, GeForce GT 630, GeForce GT 620, GeForce GT 610, GeForce 605 GeForce 500 SeriesGeForce GTX 590, GeForce GTX 580, GeForce GTX 570, GeForce GTX 560 Ti, GeForce GTX 560 SE, GeForce GTX 560, GeForce GTX 555, GeForce GTX 550 Ti, GeForce GT 545, GeForce GT 530, GeForce GT 520, GeForce 510 GeForce 400 SeriesGeForce GTX 480, GeForce GTX 470, GeForce GTX 465, GeForce GTX 460 SE v2, GeForce GTX 460 SE, GeForce GTX 460, GeForce GTS 450, GeForce GT 440, GeForce GT 430, GeForce GT 420, GeForce 405 GeForce 300 SeriesGeForce GT 340, GeForce GT 330, GeForce GT 320, GeForce 315, GeForce 310 GeForce 200 SeriesGeForce GTX 295, GeForce GTX 285, GeForce GTX 280, GeForce GTX 275, GeForce GTX 260, GeForce GTS 250, GeForce GTS 240, GeForce GT 230, GeForce GT 240, GeForce GT 220, GeForce G210, GeForce 210, GeForce 205 GeForce 100 SeriesGeForce GT 140, GeForce GT 130, GeForce GT 120, GeForce G100 GeForce 9 SeriesGeForce 9800 GX2, GeForce 9800 GTX/GTX , GeForce 9800 GT, GeForce 9600 GT, GeForce 9600 GSO, GeForce 9600 GSO 512, GeForce 9600 GS, GeForce 9500 GT, GeForce 9500 GS, GeForce 9400 GT, GeForce 9400, GeForce 9300 GS, GeForce 9300 GE, GeForce 9300 SE, GeForce 9300, GeForce 9200, GeForce 9100 GeForce 8 SeriesGeForce 8800 Ultra, GeForce 8800 GTX, GeForce 8800 GTS 512, GeForce 8800 GTS, GeForce 8800 GT, GeForce 8800 GS, GeForce 8600 GTS, GeForce 8600 GT, GeForce 8600 GS, GeForce 8500 GT, GeForce 8400 GS, GeForce 8400 SE, GeForce 8400, GeForce 8300 GS, GeForce 8300, GeForce 8200, GeForce 8200 /nForce 730a, GeForce 8100 /nForce 720

[Scionyde]

If anybody's interested, here are some quick benchmarks from Tom's Hardware.

 

http://www.tomshardware.com/news/nvidia-geforce-337.50-driver-benchmarks,26473.html

 

It's a shame they weren't able to get some SLI benchmarks, but there's some modest gains over the previous driver in Tomb Raider/BF4. The performance increase in the Star Swarm benchmark looks nice, but I'll care more about it when they look at the released game. I didn't care about the benchmark when the AMD fanboys were crapping themselves over their Mantle results before, and I won't care when Nvidia fanboys crap themselves over their 337.5 results now.

[ACatWithThumbs]

Yeah....about that...

 

Supported Products

 

ION (Desktops) ION ION LE (Desktops) ION LE GeForce 700 Series GeForce GTX TITAN Black, GeForce GTX TITAN, GeForce GTX 780 Ti, GeForce GTX 780, GeForce GTX 770, GeForce GTX 760, GeForce GTX 760 Ti (OEM), GeForce GTX 750 Ti, GeForce GTX 750, GeForce GTX 745 GeForce 600 Series GeForce GTX 690, GeForce GTX 680, GeForce GTX 670, GeForce GTX 660 Ti, GeForce GTX 660, GeForce GTX 650 Ti BOOST, GeForce GTX 650 Ti, GeForce GTX 650, GeForce GTX 645, GeForce GT 645, GeForce GT 640, GeForce GT 630, GeForce GT 620, GeForce GT 610, GeForce 605 GeForce 500 Series GeForce GTX 590, GeForce GTX 580, GeForce GTX 570, GeForce GTX 560 Ti, GeForce GTX 560 SE, GeForce GTX 560, GeForce GTX 555, GeForce GTX 550 Ti, GeForce GT 545, GeForce GT 530, GeForce GT 520, GeForce 510 GeForce 400 Series GeForce GTX 480, GeForce GTX 470, GeForce GTX 465, GeForce GTX 460 SE v2, GeForce GTX 460 SE, GeForce GTX 460, GeForce GTS 450, GeForce GT 440, GeForce GT 430, GeForce GT 420, GeForce 405 GeForce 300 Series GeForce GT 340, GeForce GT 330, GeForce GT 320, GeForce 315, GeForce 310 GeForce 200 Series GeForce GTX 295, GeForce GTX 285, GeForce GTX 280, GeForce GTX 275, GeForce GTX 260, GeForce GTS 250, GeForce GTS 240, GeForce GT 230, GeForce GT 240, GeForce GT 220, GeForce G210, GeForce 210, GeForce 205 GeForce 100 Series GeForce GT 140, GeForce GT 130, GeForce GT 120, GeForce G100 GeForce 9 Series GeForce 9800 GX2, GeForce 9800 GTX/GTX , GeForce 9800 GT, GeForce 9600 GT, GeForce 9600 GSO, GeForce 9600 GSO 512, GeForce 9600 GS, GeForce 9500 GT, GeForce 9500 GS, GeForce 9400 GT, GeForce 9400, GeForce 9300 GS, GeForce 9300 GE, GeForce 9300 SE, GeForce 9300, GeForce 9200, GeForce 9100 GeForce 8 Series GeForce 8800 Ultra, GeForce 8800 GTX, GeForce 8800 GTS 512, GeForce 8800 GTS, GeForce 8800 GT, GeForce 8800 GS, GeForce 8600 GTS, GeForce 8600 GT, GeForce 8600 GS, GeForce 8500 GT, GeForce 8400 GS, GeForce 8400 SE, GeForce 8400, GeForce 8300 GS, GeForce 8300, GeForce 8200, GeForce 8200 /nForce 730a, GeForce 8100 /nForce 720

That only means support.

That doesn't mean those cards get a boost.

In the list are also DX10 cards...

Here are some Benchmarks:

http://www.tomshardware.com/news/nvidia-geforce-337.50-driver-benchmarks,26473.html

http://wccftech.com/nvidia-geforce-337-50-beta-performance-boosting-driver-official-delivers-optimizations-directx-11-titles/

I would like to see those benchmarks with an i5.

Because the benefits should be the strongest if you have a weak CPU paired with a strong GPU.

[wtfpieisgood]

Why does sleeping dogs always get performance boosts from almost every patch it seems?

[beebskadoo]

That only means support.

That doesn't mean those cards get a boost.

In the list are also DX10 cards...

Here are some Benchmarks:

http://www.tomshardware.com/news/nvidia-geforce-337.50-driver-benchmarks,26473.html

I would like to see those benchmarks with an i5.

Because the benefits should be the strongest if you have a weak CPU paired with a strong GPU.

Nvidia never mentions the performance boosts for older cards, once the new shit comes out they support the older cards but they don't give performance gains for the old stuff. Kind of dumb.

[MayflowerElectronics]

Is there any reason this isnt being picked up my geforce experience?

 

Doesn't work on win7 64bit. That's not very smart.

[Rageodontist]

that just makes it worse

It's a Beta driver

[DaRk0]

Is there any reason this isnt being picked up my geforce experience?

 

Beta.

[rashdanml]

Linus' benchmarks lack any kind of reliability and validity, as their overclocking cannot be standardised because:

• Different GPU OC differently
• Different GPU Cores / memory within the same line of GPUs overclock differently

 

Basically all you can do with Linus' benchmarks is being happy about the colored bars.

Anyone who knows even the slightest bit about comparisons and actual measurement, can see that what he does is a waste of his time, if the goal is to be a source of information. [Ofc the goal is for you to watch the video, but lets assume he actually tried to deliver data] ^^

Agreed. Every time I watch a benchmark video, and he says "They're all OC'ed because OC'ing is easy", I just shake my head. It's also another reason why benchmark software is actually more reliable, and real-world performance isn't. It's entirely possible for two people with identical systems (identical overclocks, which is a stretch), and identical drivers/software have vastly different performance in the same game due to unexpected factors. At best, real-world benchmarks are a gamble; you may be lucky and get a good performer, or you may be terribly unlucky and get an awful performer.

 

Benchmark software is also the same test each time, whereas each reviewer has their own benchmark methodology for real-world tests, and each playthrough of said benchmark sequence can differ from test to test. 

[fojacko]

Just did a quick benchmark on the bf4 test range (not really that demanding, but i wanted to do it quickly). I ran from the US spawn to C flag taking the exact same route each time. I did this 5 times and rounded up the results.

 

335.23 - Avg: 95 - Min: 76 - Max: 123

337.50 - Avg: 97 - Min: 84 - Max: 118

 

The min frames are higher, which IMO is better than having a higher max framerate. Hell, i'm not even sure if the drivers did anything, a difference of 2 frames on average is nothing.

[Victorious Secret]

Just did a quick benchmark on the bf4 test range (not really that demanding, but i wanted to do it quickly). I ran from the US spawn to C flag taking the exact same route each time. I did this 5 times and rounded up the results.

 

 

The min frames are higher, which IMO is better than having a higher max framerate. Hell, i'm not even sure if the drivers did anything, a difference of 2 frames on average is nothing.

 

Anything that gives me a better minimum is welcome, cause that is what matters. My computer can tell me it maxed at 130 fps all day long in BF4, but if I look and see that the minimum was 55, thats shitty. If the minimum jumps to 75, well hello. Now I'm always above 60fps (and since my monitor is only 60Hz anyways) everything else is just a bonus. 

[VSG]

Agreed. Every time I watch a benchmark video, and he says "They're all OC'ed because OC'ing is easy", I just shake my head. It's also another reason why benchmark software is actually more reliable, and real-world performance isn't. It's entirely possible for two people with identical systems (identical overclocks, which is a stretch), and identical drivers/software have vastly different performance in the same game due to unexpected factors. At best, real-world benchmarks are a gamble; you may be lucky and get a good performer, or you may be terribly unlucky and get an awful performer.

Benchmark software is also the same test each time, whereas each reviewer has their own benchmark methodology for real-world tests, and each playthrough of said benchmark sequence can differ from test to test.

This plus the fact that his overclocking isn't even consistent across different cards makes me just not watch any of his benchmarks anymore. If you want to show overclocked numbers then do what enthusiasts would- overvolt, flash a 3rd party BIOS that removes some Nvidia/AMD limitations and then do a systematic core/memory/PCI-E lane overclock at the latest available WHQL and Beta driver. These days I only see benchmarks from people at stock for a new card (since they can actually be replicated and is a test for the card when you first get it). Anyway this is off topic.

I will be testing this driver (or the latest at that point) on my dual KPEs when I get back to the US in May. I will report back then.

[digitalnav]

Will be re-testing most of my cards today, SLI included.

[VSG]

Nav, can I request two sets of tests- one with a stock i3/i5 CPU and one with a 4.5GHz i7 4930k if you have one?

[digitalnav]

Nav, can I request two sets of tests- one with a stock i3/i5 CPU and one with a 4.5GHz i7 4930k if you have one?

If I get a chance. My test bench has a 4.6 4960X.

[virtualTune]

Just did a quick benchmark on the bf4 test range (not really that demanding, but i wanted to do it quickly). I ran from the US spawn to C flag taking the exact same route each time. I did this 5 times and rounded up the results.

 

 

The min frames are higher, which IMO is better than having a higher max framerate. Hell, i'm not even sure if the drivers did anything, a difference of 2 frames on average is nothing.

you are using a 660 though. The graphs the OP mentions are showing improvements for the 770 and 780 Ti specifically.

[VSG]

If I get a chance. My test bench has a 4.6 4960X.

That's of course a good replacement for the 4930k part, hopefully you get the time. Thanks man!

[Mitrailleuse]

Yeah....about that...

 

Supported Products

 

ION (Desktops) ION ION LE (Desktops) ION LE GeForce 700 Series GeForce GTX TITAN Black, GeForce GTX TITAN, GeForce GTX 780 Ti, GeForce GTX 780, GeForce GTX 770, GeForce GTX 760, GeForce GTX 760 Ti (OEM), GeForce GTX 750 Ti, GeForce GTX 750, GeForce GTX 745 GeForce 600 Series GeForce GTX 690, GeForce GTX 680, GeForce GTX 670, GeForce GTX 660 Ti, GeForce GTX 660, GeForce GTX 650 Ti BOOST, GeForce GTX 650 Ti, GeForce GTX 650, GeForce GTX 645, GeForce GT 645, GeForce GT 640, GeForce GT 630, GeForce GT 620, GeForce GT 610, GeForce 605 GeForce 500 Series GeForce GTX 590, GeForce GTX 580, GeForce GTX 570, GeForce GTX 560 Ti, GeForce GTX 560 SE, GeForce GTX 560, GeForce GTX 555, GeForce GTX 550 Ti, GeForce GT 545, GeForce GT 530, GeForce GT 520, GeForce 510 GeForce 400 Series GeForce GTX 480, GeForce GTX 470, GeForce GTX 465, GeForce GTX 460 SE v2, GeForce GTX 460 SE, GeForce GTX 460, GeForce GTS 450, GeForce GT 440, GeForce GT 430, GeForce GT 420, GeForce 405 GeForce 300 Series GeForce GT 340, GeForce GT 330, GeForce GT 320, GeForce 315, GeForce 310 GeForce 200 Series GeForce GTX 295, GeForce GTX 285, GeForce GTX 280, GeForce GTX 275, GeForce GTX 260, GeForce GTS 250, GeForce GTS 240, GeForce GT 230, GeForce GT 240, GeForce GT 220, GeForce G210, GeForce 210, GeForce 205 GeForce 100 Series GeForce GT 140, GeForce GT 130, GeForce GT 120, GeForce G100 GeForce 9 Series GeForce 9800 GX2, GeForce 9800 GTX/GTX , GeForce 9800 GT, GeForce 9600 GT, GeForce 9600 GSO, GeForce 9600 GSO 512, GeForce 9600 GS, GeForce 9500 GT, GeForce 9500 GS, GeForce 9400 GT, GeForce 9400, GeForce 9300 GS, GeForce 9300 GE, GeForce 9300 SE, GeForce 9300, GeForce 9200, GeForce 9100 GeForce 8 Series GeForce 8800 Ultra, GeForce 8800 GTX, GeForce 8800 GTS 512, GeForce 8800 GTS, GeForce 8800 GT, GeForce 8800 GS, GeForce 8600 GTS, GeForce 8600 GT, GeForce 8600 GS, GeForce 8500 GT, GeForce 8400 GS, GeForce 8400 SE, GeForce 8400, GeForce 8300 GS, GeForce 8300, GeForce 8200, GeForce 8200 /nForce 730a, GeForce 8100 /nForce 720

the update supports them,but they don't get the boost that the 6XX and 7XX gets.

[nokturrduk84]

Nvidia 4 life!

The only true GPU leader

[pixeldensity]

 

Nvidia 4 life!

The only true GPU leader

 

wow that all you have contribute to the conversation. Sad

 

On another note, so I took a look at the test TH did and the number gains don't look that impressive however, maybe they should of use a lower end cpu just like mantle sees large gains with lower end cpus.