What happened to the nvidia drivers? What happened to us? Activism time!

Discussion in 'Gaming (Software and Graphics Cards)' started by Huxlay, Feb 22, 2016.

Thread Status:
Not open for further replies.
  1. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,866
    Trophy Points:
    931
    We're talking mobile here, not desktop. Titan X is a desktop card, and there, you have AMD options. On laptops there are none. Maybe one or two with R9 M380 which is barely equivalent to a 128-bit GTX 965m. Otherwise you're looking at an R9 M270 or M275 which doesn't even compete with a 950m. You really have to be a die hard Nvidia hater to deny yourself that level of performance difference especially if you are a gamer that likes to play the latest AAA games.

    I mean if they were killing puppies because I was buying Nvidia, then yeah, I'd probably suffer with AMD GPU and the slim amount of laptop offerings. But it's not anything like that.
     
  2. sasuke256

    sasuke256 Notebook Deity

    Reputations:
    495
    Messages:
    1,440
    Likes Received:
    449
    Trophy Points:
    101
    Plus the power consumption is too high, I still don"t know if Enduro Tech is correctly working, for laptop, AMD has nothing to compete with Maxwell in terms of performance/watt ratio..
     
    transphasic and HTWingNut like this.
  3. iaTa

    iaTa Do Not Feed

    Reputations:
    1,328
    Messages:
    2,675
    Likes Received:
    197
    Trophy Points:
    81
    Does the GTX 980 for notebooks suffer from the same issues?
     
  4. Arog

    Arog Notebook Consultant

    Reputations:
    15
    Messages:
    252
    Likes Received:
    44
    Trophy Points:
    41
    I use the latest nvidia drivers, what is wrong with them? Sad to see AMD unable to compete, and with nvidia products reasonably priced there is zero reason to go with AMD. I do hope AMD gets back in the game and makes things interesting.


    Sent from my iPhone using Tapatalk
     
  5. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,866
    Trophy Points:
    931
    What issues?
     
  6. pete962

    pete962 Notebook Evangelist

    Reputations:
    125
    Messages:
    498
    Likes Received:
    222
    Trophy Points:
    56
    I think he was refering to high power usage by AMD, so was the same issue present with desktop GTX?
     
  7. iaTa

    iaTa Do Not Feed

    Reputations:
    1,328
    Messages:
    2,675
    Likes Received:
    197
    Trophy Points:
    81
    Sorry, I should have quoted:

     
  8. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,866
    Trophy Points:
    931
    Ah yes. It's still considered a mobile GPU. Just mobile GTX 980 not desktop GTX 980 even though it's same silicon.
     
    D2 Ultima likes this.
  9. Huxlay

    Huxlay Notebook Consultant

    Reputations:
    12
    Messages:
    131
    Likes Received:
    69
    Trophy Points:
    41
    @Mr. Fox, would you kindly provide me with a list of all the issues and benchmarks that you have made on the drivers over the year? I'll try to make our case to the PC Master Race community, if they see how ****ty the hand was dealt to us, perhaps they will help us get another twitter **** storm up and running.
     
  10. nipsen

    nipsen Notebook Ditty

    Reputations:
    694
    Messages:
    1,686
    Likes Received:
    131
    Trophy Points:
    81
    I did enjoy the entire read, and so on. Entire thread, really. And don't want to rain on anyone's parade-march.

    But the truth is that graphics cards, peripheral external cards, essentially peaked out a very long time ago. And what actually stalled the mobile and frankly the desktop space, and gave us what we have now - was that AMD was cut off from releasing the apu project they were planning, in the form it actually would have worked in, by Intel lawsuit. Specifically that they were required to separate the cpu and gpu interfaces on the hardware interface level. Not only that, Nvidia was cut off from making a similar offering with ION prototype - that then eventually was released with a discrete intel graphics module rather than what was planned.

    Intel has then continued to release their usual incremental updates as they have done for many years. And Nvidia has geared itself into that by releasing more and more energy-efficient cards in the medium to small (consoles to laptops) area, and matched their business with the existing model. It's a good business-move by both intel and nvidia, but we of course suffer as customers - the longer it takes for "OpenCL-type" based computers with generic processing units that can be programmed to run various different instruction sets, outside the current ISA, with the required sandwich of components that obviously could have (and have, in protoypes) been removed. And replaced with an integrated bus.

    Rather than what we have now when people seriously are suggesting that what we need is a faster interconnected bus that can program it's own IO, etc. Anything to keep the current industry standard(ISA) must be done, see, or else large parts of the hardware business collapses.

    So the thing is that no matter how much you pressure Nvidia, they're not going to start making stuff like ION with a graphics module integrated on the cpu in form of several generic processing units. And they cannot magick out performance increases on the driver level at the stage where post-processing filters draw about 90% of the actual graphics card grunt. Instead you get mobile and laptop drivers that have the biggest % increases, along with a hilarious subset of overclocks in an attempt to stretch the tdp budget. None of this is going to produce the kind of results we saw where most of the driver optimisations were culling and overdraw routines. In the same way, most of the OpenCL-type routines are not even used in any common sdk, meaning that the driver-optimisations on queuing and multitasking, upgrades for cards that deal with such, etc., are a waste of money for gamers. That's just how it is, and how it's going to be, until .. I don't know.. Intel goes out of business, and no one takes up their torch in maintaining the "cpu as part of our motherboard sandwich, or else it's illegal" paradigm.

    Or at the point where ARM manages to push intel out of the market (haha, yeah, I know), without getting sabotaged to death. They could offer certain mobile laptop solutions at the moment, but they are not part of the laptop market or desktop market, very likely for purely business-diversification reasons.

    Or possibly at the point when some explicitly parallel assembly language is adopted widely and entirely new hardware will be manufactured to be compatible with each other as long as they use the interfaces required by the interconnected bus standard that would go along with it.

    But that simply won't happen in our lifetimes. In fact, the only commercially available product with an explicitly parallel assembly language running on several generic computing cores - was dropped and removed from marketability again. The ram-solution with concurrent reads and writes in that product was dropped in favor of one that adopts a solution for a higher frequency exclusive access writes and reads. That never is going to come near the potential of that dropped solution. That's the long and the short of it. The most we can hope for is semi-integrated components floating up from a scuttled AMD frigate that does away with parts of the sandwich-board, while integrating certain components of the gpu on the same die as the cpu. Nothing beyond that will happen in the "PC" space as long as Intel exists, and as long as there is no explicitly parallel assembly language widely adopted by all major interests in the industry. And that will not happen as long as there's money to be made on incremental updates. Simple as that.
     
    transphasic likes this.
Loading...
Thread Status:
Not open for further replies.

Share This Page