Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

How long will a GTX 970 last

1235»

Comments

  • Leon1eLeon1e Member UncommonPosts: 791
    edited April 2016
    Torval said:
    Leon1e said:
    Hmm you should really wait. Last I heard the 9xx series have architectural limitation when it comes to DirectX12/Vulkan and most nVidia fans will cry out in horror once these games start to pop out. It was something with AsyncCompute, last I heard they were emulating it and thus lowering FPS. I'm sure there are far more knowledgeable people than me to confirm/deny this but it's literally in almost every nvidia-vs-amd benchmark topic. Currently nVidia has a worse driver.
    The OP wants to use Shadowplay so AMD isn't an option. For 1080p gaming those Vulkan/Dx12 issues at 1440 or 4k aren't going to be a factor.

    Unless Shadowplay requires a huge chunk of video memory the 970 should be adequate. The 3.5GB effective memory of the 970 is probably one of the most glaring weaknesses. It hasn't really been a factor for me in all but a couple of games.
    Are you implying that there is no video capture software, like say, OBSProject? That rivals shadowplay in performance (if not beating it). Tying a hardware purchase to a software product that isn't even a game is silly man ...

    Also, here's some Dx 12 performance charts http://www.extremetech.com/gaming/226082-early-directx-12-games-show-a-distinct-split-between-amd-nvidia-performance?sf24015547=1

    A lot of developers are working with nVidia to fix that broken Async Compute on nvidia hardware but It's not an easy task if the underlying architecture is not asynchronous

    EDIT: Actually AMD does have a shadowplay-like client. It's basically Raptr in AMD skin but it checks for drivers, game optimizations and whatnot and yes it has game capture too. 
  • SEANMCADSEANMCAD Member EpicPosts: 16,775
    edited April 2016
    SEANMCAD said:
    @SEANMCAD  This is relatively off topic but yes, Moore's Law may be coming to an end. Just because it's a "law" doesn't mean it is a law by scientific standards.

    Several articles on this both scholarly and un-scholarly. A quick google search of "the end of moore's law" shows this.

    I will point out that the law does stand in one notion: size. While chips may not be getting "faster" or more "capable" they are certainly getting smaller, and that's about it. Complicated as well however. Manufacturers have to make these things for broad consumer mass. By technicality irrelevant to chip size it is extremely possible to manufacture any component at over 10 times that of the current zenith consumer parts. 
    but if Moores Law is over is it over in the current iteration of manufacturing because if it is not then one can expect the performance and price to performance to be the same % change as it was last generation and generation before that and etc.

    Expected yes. Throw free-market economics into that mix and the craziness begins of course.

     
    hold on there cowboy!
     1. we dont have a 'free-market' economy
    2. moores law has been 100% accurate every iteration over the past 30 years or so given the same so called 'free-market' we are in now. so what specfically would be different here in which moores law would be 'trumped'?

    Please do not respond to me, even if I ask you a question, its rhetorical.

    Please do not respond to me

Sign In or Register to comment.