New nvidia beta driver owns AMD again in BF3 and MOHWF.

Just installed them.About 15 fps boost in BF3SP same goes to MOHWF about 12fps boost.Also skyrim got about 8fps boost.Real nice drivers. :D

http://www.geforce.com/whats-new/articles/nvidia-geforce-310-54-beta-drivers-released
51 answers Last reply
More about nvidia beta driver owns mohwf
  1. What card?
  2. @cooldudesubho
    is it only for 680s and 660s?
    I've benched witcher 2, ran unigine heaven on my gtx 660 ti but didn't notice an increase or decrease in performance.I used perform clean install while installing drivers btw.
  3. No its for every card. I've noticed positive increase in performance so far int he games I've tested. Me and cool have the same card.

    They get my stability approval so far. and Recommended.

    NVIDIA: 1
    AMD: 0


    Game, Set, Match
  4. GTX 670.I got boost in skyrim,BF3,MOHWF,sleeping dogs.

    I am yet to test unigine.And witcher2 is POS.I don't play.
  5. AMD will benefit in Witcher 2 and MOH:WF because they helped in development and partnering. Which is why those games are always easy to run..
  6. @rockdpm
    on overclocked or on stock clocks?
    @cool
    how much increase did u get on sleeping dogs and at wat settings?
  7. my card had a slight overclock and along with boost. i see a max frequencyof 1100mhz. see people look at drivers only for the games. I also judge a driver by the stability it provides to the chip and how much more i can achieve out of the card.
  8. Rockdpm said:
    AMD will benefit in Witcher 2 and MOH:WF because they helped in development and partnering. Which is why those games are always easy to run..


    the Witcher 2 is TWIMTBP title. so there are performance increase with this new beta driver? i'm still on 306.97
  9. renz496 said:
    the Witcher 2 is TWIMTBP title. so there are performance increase with this new beta driver? i'm still on 306.97

    strange... it says gaming evolved on opening witcher 2.
    I OC'ed my msi 660 ti from 1020 base to 1100 base.I see maximum boost clocks of 1267.Hope this doesn't affect the performance boost that i could've got from the drivers.
  10. AMD just updated their Catalyst 12.11 towards Beta 7


    10%-15% more performance in Battlefield 3 in most cases

    More than 20% in certain missions and sequences (Comrades)

    Up to 7% more performance in Metro 2033
    Up to 10% more performance in DIRT Showdown
    Up to 8% more performance in Sleeping Dogs
    Up to 12% more performance in Civilization V
    Up to 10% more performance in StarCraft II
    Up to 8% more performance in Sniper Elite: V2
  11. The driver *wars* have gotten insane (in a good way from the looks of them) these last few weeks.
  12. Anik8 said:
    @rockdpm
    on overclocked or on stock clocks?
    @cool
    how much increase did u get on sleeping dogs and at wat settings?

    High AA,everyhting else max.I didn't play much.But it seems it helped to reduce framerate drop while driving during daytime.1600-900 res.I will play it more then report.Even max pyne3 got a boost.About 2 to 3 frames.
  13. cooldudesubho said:
    High AA,everyhting else max.I didn't play much.But it seems it helped to reduce framerate drop while driving during daytime.1600-900 res.I will play it more then report.Even max pyne3 got a boost.About 2 to 3 frames.

    On High AA my min FPS increased by 5 and on extreme AA by 3.That's the same I get with my OC.But the avg fps remains the same.This is from the in game benchmark.
    Now I get +10 min fps in High AA(drivers + OC) and +6(3+3) in extreme AA.So yes +1 to smoothness.But the avg fps hasn't increased.On 306.97 my avg fps were 34.7(extreme AA)/69.8(high AA) at stock and 38.7(extreme AA)/75.6(high AA) after OC.After 310.54 I get the same figures.
    About the min fps I've seen improvements all over,even in heaven 3.0 from 15 fps to 20 fps.
    3dmark11 scores are a bit lame X3138 vs earlier X3131,(+8 to gfx) score and P8879 vs earlier P8875,(+20 to gfx score).
  14. I see no "ownage" here. Both teams are just continuing with incremental improvements. In general, whether you choose red or green for your graphics card, you'll be able to enjoy your games just fine.
  15. Anik8 said:
    On High AA my min FPS increased by 5 and on extreme AA by 3.That's the same I get with my OC.But the avg fps remains the same.This is from the in game benchmark.
    Now I get +10 min fps in High AA(drivers + OC) and +6(3+3) in extreme AA.So yes +1 to smoothness.But the avg fps hasn't increased.On 306.97 my avg fps were 34.7(extreme AA)/69.8(high AA) at stock and 38.7(extreme AA)/75.6(high AA) after OC.After 310.54 I get the same figures.
    About the min fps I've seen improvements all over,even in heaven 3.0 from 15 fps to 20 fps.
    3dmark11 scores are a bit lame X3138 vs earlier X3131,(+8 to gfx) score and P8879 vs earlier P8875,(+20 to gfx score).

    How your 3dmark score higher than mine gtx 670?I get p8820 all stock.What you stock score?How much average you get in sleeping dogs benchmark at 1600-900res high AA.
  16. I've OC'd my msi 660 Ti power edition card @1100(base)-1267(max) on core and 1777(7.1GHz) on memory.
    Core voltage: +25mv
    Power Limit: 114%
    Core clock: +80
    Memory clock: +550
    Fan speed: 70%(Manual)
  17. I have not tested sleeping dogs @ 1600/900.However my stock score on sleeping dogs @ 1080p is given above.
    On 3Dmark I get ~8100 at stock.I don't remember the exact number but even at stock my card was factory oc'ed by 105 mhz on the base clock by msi.
  18. Ill run some test later this afternoon. And will post results
  19. Anik8 said:
    I have not tested sleeping dogs @ 1600/900.However my stock score on sleeping dogs @ 1080p is given above.
    On 3Dmark I get ~8100 at stock.I don't remember the exact number but even at stock my card was factory oc'ed by 105 mhz on the base clock by msi.

    Can you run at that res.Just curious.With and without OC.
  20. Rockdpm said:
    No its for every card. I've noticed positive increase in performance so far int he games I've tested. Me and cool have the same card.

    They get my stability approval so far. and Recommended.

    NVIDIA: 1
    AMD: 0


    Game, Set, Match

    I kind of wonder what the last ATI/AMD card you had was? I have not seen an ATI driver crash since the 9800Pro days.
    While I did have a full 9 months of daily crashes with my 8800GTX(yeah near 600 dollar card and that long for them to get it working in Vista). This was AFTER SP1 was out. Was ready to just go back to XP but decided to try my old X1900XT and it worked(and after they finally got the drivers sorted, the 8800GTX worked too). Sure it was slow, but it worked fine for the next 9 months while I tried each beta and official driver along the way to try to get that 8800GTX working. Nvidia told me it was a regression in the drivers.

    I am hearing some horror stories on the 7XXX cards, but have not tested any to see what it is like.

    This is not to slam Nvidia or anything. I am looking at a new 6XX card soon.
  21. Nvidia has had almost as many driver issues as AMD lately and that's still not much for either company. Since Catalyst 12.6 and 12.7 several months ago now, AMD's drivers have been top-notch with Nvidia and it's not like Nvidia didn't have major hiccups of their own with the 6xx card's drivers earlier on too. Many Nvidia card owners had severe stuttering, V-Sync, and GPU frequency throttling problems for a while. Right now, they're both doing great with the drivers.
  22. oO, TXAA for Black Ops II and Assassins Creed III!!
  23. cooldudesubho said:
    Can you run at that res.Just curious.With and without OC.

    Sleeping Dogs stock/OC @ 1600/900
    AA extreme,rest max motion blur high - avg 48.5/54.1, max 64.8/83.2, min 31.8/35.5, gpu usage ~80%/~90%
    AA high,rest max motion blur high - avg 88.6/94.2, max 132.2/138.4, min 52.1/58.0, gpu usage ~70%/~80%

    At 1920x1080p I always get gpu usage above 90%
  24. The true story behind NVidia's R310 drivers are that they support Linux. With NVidia AND Valve actively working on Linux, you can say goodbye to bloatware-ridden Windows... if you want to, of course. I do.
  25. I wonder if they (both companies) sacrifice image quality to get these kind of performance leap. These are just too high of an improvement to get with simple driver optimizations imo
  26. Anik8 said:
    Sleeping Dogs stock/OC @ 1600/900
    AA extreme,rest max motion blur high - avg 48.5/54.1, max 64.8/83.2, min 31.8/35.5, gpu usage ~80%/~90%
    AA high,rest max motion blur high - avg 88.6/94.2, max 132.2/138.4, min 52.1/58.0, gpu usage ~70%/~80%

    At 1920x1080p I always get gpu usage above 90%

    Hmm i get 91.8.That about right since your stock is also OC.Vaniila will get another 5 fps less.My min is 57BTW.
  27. You may be right. Or you may just think that the new architecture isn't optimized yet. By February on the release of the 8970 and the late release of the 780 you shouldn't see these increases because they should've optimized the 28nm new process completely :)
  28. I don't think image quality is being sacrificed.AMD has GCN to tweak with while NVIDIA has CUDA 5 with which they can make use of dynamic parallelism and GPU direct to boost kepler cards.So I think this driver war is quite legit IMO.
  29. i8myhippo said:
    The true story behind NVidia's R310 drivers are that they support Linux. With NVidia AND Valve actively working on Linux, you can say goodbye to bloatware-ridden Windows... if you want to, of course. I do.


    None of that matters if most games are made with DirectX and Linux doesn't have DirectX support. Steam working on Linux has nothing to do with games working on Linux.
  30. hapkido said:
    None of that matters if most games are made with DirectX and Linux doesn't have DirectX support. Steam working on Linux has nothing to do with games working on Linux.


    Through WINE and such, Linux has DX support at least up to 10.1, IDK if they have DX11/11.1 working yet. Many modern games can work on Linux. For example, Crysis 2, WoW, and much more run great on Linux.
  31. Onus said:
    I see no "ownage" here. Both teams are just continuing with incremental improvements. In general, whether you choose red or green for your graphics card, you'll be able to enjoy your games just fine.



    Agreed. And pricing still favors AMD. Nvidia still needs that $50-70 price drop and a game bundle.
  32. ^ honestly i doubt that nvidia will drop their price much because even with their current deal (more expansive, less game bundle) they still able to sell their card just fine
  33. Current Pricing

    NVIDIA GTX 680 – A reference build is currently available for $459.99 with factory overclocked offerings starting at $469.99

    NVIDIA GTX 670 – A reference build is currently available for $359.99 with factory overclocked offerings starting at $389.99. There are several other factory overclocked cards that start lower than $389.99, however, the factory overclock on those cards are so small it is not going to make a meaningful performance difference.

    NVIDIA GTX 660 Ti – A reference build is currently available for $289.99 with factory overclocked cards that start at $299.99.

    AMD HD 7970 GHz Edition – A factory overclocked build is currently available for $449.99, which is priced lower than the reference clocked HD 7970 GHz Edition cards that are currently available.

    AMD HD 7950 Boost Edition – A reference build is currently available for $319.99.

    AMD HD 7870 GHz Edition – A reference build is currently available for $229.99 and a factory overclocked version is available for $239.99.
  34. Anik, except the 680 is comparable to 7970 ghz, 670 to 7970 stock, 660 ti to 7950, and 660 to 7870.
  35. Not quite true anymore. AMD has the 7870 more comparable to the 660 Ti than to the 660 and the 7970 is more comparable to the 680 with the 7970 GHz Edition coming out a little on top. The 7950 Boost isn't really ahead of the 670, but it's right up there with it.

    Of corse, excluding the 660, 660 Ti, and 7870, these other cards are all mostly so close (at reference) that it's not like it makes a significant difference either way in most situations, but still.
  36. @Ironslice
    Yeah it seems like that.However I got this from [H]'s latest article on beta drivers made on 12.11.12.So I can't tell if its from newegg/bestbuy/amazon.
  37. @blazorthon, Ironslice was talking about the price not the performance.And even if you do look at performance you must look at it with 12.11 vs 310.54 not any earlier versions.Unless a thorough review is done by a well known website we can't jump to conclusion regarding performance.
  38. There are seven versions of 12.11 and several of them have performance that varies almost as much as the first version did from Catalyst 12.10. As 310.54 came out, AMD also threw out another revision of 12.11 and reclaimed the ground that AMD lost from 310.54. Sorry if I missed Ironslice's point assuming that it was about price rather than performance, but what I said still stands in the context of performance. Sure, it's not an absolute truth for all games, but it is a general average.
  39. The truth is that AMD just abandoned there old series card. 6970 got zero improvemnt in this new drivers.While still now fermi gets decent boost with all new driver. Proof posted by recon_UK.

    http://battlelog.battlefield.com/bf3/forum/threadview/2832654625316638963/

    That is another reason why people buy nvidia card even at premium.Longer support.Just like apple.Even old devices get latest version of software.I bet samsung will never do that.lol
  40. As far as I've heard the latest revision of 12.11 only benefits crossfire users not single card users.
    Honestly I expected amd to release these kind of performance drivers several months ago.What took thwm so long IDK.Till the end of october nvidia were busy releasing drivers to support their newest members of the kepler family and this november when those issues has been solved, they've gone full swing into releasing performance oriented drivers.
    I bet,from now on and till 2014 we'll keep noticing newer and newer drivers resulting in switch in performance charts from time to time.But that doesn't mean people should recommend cards based on performance boost from drivers as it is temporary and subject to change.Cards should be chosen according to their performance at launch as both the architectures have similar potential to get boosted from drivers.
    The only thing that stands out is that nvidia has joined the party late.Suppose amd did 7 or 8hrs ago while nvidia did 2hrs ago.
  41. AMD did release performance improvement drivers several months ago. Why aren't you counting Catalyst 12.7 as a good performance-improving driver?
  42. blazorthon said:
    AMD did release performance improvement drivers several months ago. Why aren't you counting Catalyst 12.7 as a good performance-improving driver?

    It only helped 7000 series.lol
  43. Yes they did but still the performance gap didn't marginalize because of that 25% and it took them 5 more months to do so with a driver update that enabled 7xxx cards almost 90% boost in one or two games.
    Suppose nvidia has joined the party now with 15% and 26% initial driver boosts from their part this month onwards.Get my point? blazorthon?
    Just as GCN was much newer in June,Cuda 5 is even younger now which comes with several features that can boost performance by quite an extent take dynamic parallelism for example.
  44. Anik8 said:
    Yes they did but still the performance gap didn't marginalize because of that 25% and it took them 5 more months to do so with a driver update that enabled 7xxx cards almost 90% boost in one or two games.
    Suppose nvidia has joined the party now with 15% and 26% initial driver boosts from their part this month onwards.Get my point? blazorthon?
    Just as GCN was much newer in June,Cuda 5 is even younger now which comes with several features that can boost performance by quite an extent take dynamic parallelism for example.

    In some peoples eye's AMD/ATi will always be better regardless of their failings, the same could be said of Nvidia for others as well of course so the point is that there is no point in arguing just move on.
  45. Onus said:
    I see no "ownage" here. Both teams are just continuing with incremental improvements. In general, whether you choose red or green for your graphics card, you'll be able to enjoy your games just fine.


    Agreed.....right now it's just about what ya gonna pay.....the driver upgrades shouldn't be changing ya mind between a 670 and 7970.

    This is no different from the barrage of recent posts about the Catalyst 12.11 drivers and the Techpowerup review that had the 7970 leapfrogging over 670 the when the new 12.11 AMD drivers were compared against the 670 w/ nVidia's 2 generations old drivers.

    That techpowerup article had the 670 going from 4% faster than the 7970when both had older drivers to 3 % slower ..... but again with nVidia running 2 generation old drivers and AMD running the latest and greatest. NOTE: I am not condemning the author of this article as I believe his intent was to show the difference between the old and new catalyst drivers. It's the unsupported conclusions drawn by readers that the new AMD / old nVidia performance results are to somehow have meaning in comparing what one can buy today that I take issue with.

    The same applies to his new claim with the 310.54 drivers ..... meaningless if ya not comparing with the latest and greatest AMD drivers.

    But does it even matter ? Not with current pricing. Best "card" deal at the moment is easily the factory overclocked 670 from MSI. The 7970 card / game bundle could is a decent deal if ya want those 3 games and the cost of those might be worth a large part of that $77 cash difference. Otherwise, even ignoring the old nVidia drivers used in the test and the fact that the 670 below is factory OC'd and guaranteed at that speed, the 3% more performance difference between reference cards doesn't warrant a 25+% cost increase.

    $313 factory OC'd GTX 670 ($370 - 10%off promo - $20 MIR)
    http://www.newegg.com/Product/Product.aspx?Item=N82E16814127675

    $390 reference MSI HD7970 ($420 - $30 MIR)
    http://www.newegg.com/Product/Product.aspx?Item=N82E16814127645

    When I can buy a comparable non reference HD7970 for 3 -5 % more cash from any major vendor (Asus, Gigabyte, MSI, EVGA, MSI) than I can buy a non reference GTX 670 from the same major vendors, I'll go back to calling the choice a toss up, assuming of course that a performance comparison with both cards using latest drivers or significant price shifts change my mind in one direction or another. Right now I have $77 worth of reasons to pick the 670.

    Getting back to driver leapfrogging .....Yes, each driver release will correct some buglets, resolve some compatibility issues but more importantly this "competition" has resulted in a tweaking exercise from both camps to "get better benchmarks. How much of these "driver optimizations" are legit improvements and how much is actual "cheating" (decrease of precision of render targets, textures, quality decreases and other tricks) remains to be determined.....tho in some cases it has been actually detected and reported.

    http://benchmarkreviews.com/index.php?option=com_content&task=view&id=12845&Itemid=47

    As the author said, one has to question the legitimacy when the article is written by the competition, but it was independent review sites which provided the basic information. It would be nice if we didn't have to rely on translations if german sites to get this kind of info. And by no means am I suggesting that nVidia doesn't pull a few tricks of their own in their driver updates.

    But when I see comparisons between one set of cards with new drivers, versus another set of cards w/ 2 generation old drivers and this data being used as the basis of recommendations, I have to raise the "buyer beware" flag. The 7970 and 670 could continue to leap frog over one another for months to come but I'd be hesitant to draw conclusions about the effect of driver updates when:

    -The sources quoted use new drivers for one manufacturer and old drivers for the other.

    -The sources quoted have not tested or looked for any changes in how the drivers run the benchmarks with regard to reductions in quality or settings. Yeah, you might not notice a difference in quality with a driver "optimization" that drops quality to go from 70 to 80 fps. But to claim some kind of "victory" over a card that gets 75 fps w/o that quality reduction is simply not a legitimate comparison.
  46. ^ while I agree with what your saying, you have to remember pricing is a bit dynamic just looking at new egg I found this HIS 7970 card for $399 plus 30 rebate so $369.99 Plus you get Three games free.

    http://www.newegg.com/Product/Product.aspx?Item=N82E16814161412

    So the 670 maybe save you $57 until you include the free games assuming you want them that pretty much erase the price difference. Once both rebate discounts stop then you only $20 difference in full retail costs.
  47. nukemaster said:
    I kind of wonder what the last ATI/AMD card you had was? I have not seen an ATI driver crash since the 9800Pro days.
    While I did have a full 9 months of daily crashes with my 8800GTX(yeah near 600 dollar card and that long for them to get it working in Vista). This was AFTER SP1 was out. Was ready to just go back to XP but decided to try my old X1900XT and it worked(and after they finally got the drivers sorted, the 8800GTX worked too). Sure it was slow, but it worked fine for the next 9 months while I tried each beta and official driver along the way to try to get that 8800GTX working. Nvidia told me it was a regression in the drivers.

    I am hearing some horror stories on the 7XXX cards, but have not tested any to see what it is like.

    This is not to slam Nvidia or anything. I am looking at a new 6XX card soon.

    Did i say ANYTHING about AMD's drivers being unstable ANYWHERE in that post. Actually for the record. DID i say ANYTHING about AMD drivers in that post period? No. My point was in performance the new NVIDIA drivers met my approval. And don't worry about what cards I have or have had. Not of your concern
  48. I completely agree with you JackNaylorPE,that was quite what I was trying say not only replying to blazorthon here but in several other posts getting wild,most probably where the author asks about a 670 or 7970.
    Thanks mousemonkey, for pointing it out.Sometimes I do get carried away when some people throw statements backed by fabricated references looking at which I wonder if my 660 ti has some magical power so as to score 20-25 fps higher than what their charts show at same settings.
    Even I read that article by [H] where the 660 ti and 7870 parity surfaced.Trust me I can immediately make a record of the sleeping dogs benchmark where I get 20 fps more at the same settings.
  49. Here is some benchmarks from a friend over on Battlelog.

    I will directly copy and paste it here for you all.




    Here are my results for a single EVGA GTX 480 SSC.


    Test setup



    i5 2500k 4500mhz
    EVGA Z68 SLi motherboard
    8GB Gskill Ripjaws X 1600mhz
    EVGA GeForce GTX 480 SSC (850/1700/2000) (Nvidia 310.54 BETA)
    2x Samsung spinpoint F3 1TB RAID-0
    Silverstone Strider 1000w 80 Plus Gold

    None recorded benchmark (full potential)

    Frames, Time (ms), Min, Max, Avg
    14089, 234797, 32, 147, 60.005

    That is an insane improvement!







    Recorded benchmark

    Tested maps
    Gulf of oman-Conquest
    Noshahr canals-TDM
    Damavand peak-Conquest

Ask a new question

Read More

Nvidia Battlefield AMD Graphics Product