Pascal: What do we know? Discussion, Latest News & Updates: 1000M Series GPU's

Discussion in 'Gaming (Software and Graphics Cards)' started by J.Dre, Oct 11, 2014.

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

    moviemarketing Milk Drinker

    Reputations:
    1,036
    Messages:
    4,256
    Likes Received:
    881
    Trophy Points:
    181
    Do you imagine Nvidia will repeat the (desktop) GTX 980 for mobile experiment with Pascal's GP204 or whatever it will be called? If so, is that likely to be another year or two away?

    Otherwise, the full GTX 980 with 2048 CUDA cores in P870DM for example is still going to be faster than the 1080M for the foreseeable future, no?
     
  2. J.Dre

    J.Dre Notebook Nobel Laureate

    Reputations:
    3,700
    Messages:
    8,316
    Likes Received:
    3,819
    Trophy Points:
    431
    The 980M is about 30% less powerful than the 980. The 1080M will definitely be more powerful than a single 980.

    From what I recall when they made the 980M, NVIDIA wanted to close the gap between desktop and mobile graphics performance, so seeing another 1080 "mobile version" is very likely, IMO.
     
    moviemarketing likes this.
  3. Bankfenster

    Bankfenster Notebook Enthusiast

    Reputations:
    0
    Messages:
    37
    Likes Received:
    1
    Trophy Points:
    16
    Hi,
    I got same questions:
    - Do you think there is a significant performance improve in "casual" gpus like 1040m / 1050m?
    - I read a lot, that pascal could be hot, but why? If they improve the performance per watt, why should pascal be hotter than maxwell? For example, when I compare the tests of notebooks (notebookcheck.net), kepler GPUs run at max like 10°C cooler than fermi GPUs (while having more or less the same cooling system).
     
  4. Cloudfire

    Cloudfire (Really odd person)

    Reputations:
    7,279
    Messages:
    10,310
    Likes Received:
    2,878
    Trophy Points:
    581
    Who knows what the performance of GP104/1080M will be.
    I would guess a bit faster than GTX 980 but I do not expect a lot faster. For one the GTX 980M is a massive chip, it is 400mm2 and I think Nvidia would want to go down to 350mm2 again which these chips usually have been.
    Secondly, I think it will be a 100W GPU which means it get a disadvantage against the 165W GTX 980 we now have.

    If I should guess, 40-60% faster than GTX 980M and 0-20% faster than notebook GTX 980.
    If its another 165W GPU, that of course changes things. But I think Nvidia will stick with 100W with disabled cores first, then followed up by a full GP104 chip later on in 2017 which is either a 105W GPU if they are successful at reducing voltage with the Pascal silicon so they can add more cores without going up in wattage. Or it is another 165W GPU.

    GTX 680M was about 40-50% faster in average than GTX 580M when it launched. It was bigger, and was based on a new architecture and a new node. I thought it was almost twice as fast, but these links says otherwise
    http://www.notebookcheck.net/Review-GeForce-GTX-680M-vs-Radeon-HD-7970M.77110.0.html
    http://www.anandtech.com/show/6336/...book-review-nvidias-geforce-gtx-680m-in-sli/3
     
    Last edited: Mar 5, 2016
    Kaozm, Robbo99999 and moviemarketing like this.
  5. HTWingNut

    HTWingNut Potato

    Reputations:
    21,578
    Messages:
    35,409
    Likes Received:
    9,863
    Trophy Points:
    931
    980M is really a 120W TDP MXM GPU, and if they make equivalent Pascal TDP, I would expect nearly double performance from the die shrink and architecture changes. If they drop to a 100W TDP well, then it's not really an apples to apples comparison. Would be a bit disappointing to say the least. Depends on what vRAM is used as well, etc.
     
    Robbo99999 likes this.
  6. tgipier

    tgipier Notebook Deity

    Reputations:
    197
    Messages:
    1,596
    Likes Received:
    1,572
    Trophy Points:
    181
    LOL. Isnt it just a m5000m card?
     
  7. n=1

    n=1 YEAH SCIENCE!

    Reputations:
    2,544
    Messages:
    4,346
    Likes Received:
    2,600
    Trophy Points:
    231
    One thing's for sure: don't expect 100% GP104 potential out the gate. We were stuck on 28nm for 4 years (and counting), and we'll be on 16/14nm for even longer, so it's gonna be a loooooooooooooooooong milking process.
     
    Mr Najsman, jaybee83 and Prema like this.
  8. Splintah

    Splintah Notebook Deity

    Reputations:
    278
    Messages:
    1,948
    Likes Received:
    595
    Trophy Points:
    131
    [​IMG]
     
  9. J.Dre

    J.Dre Notebook Nobel Laureate

    Reputations:
    3,700
    Messages:
    8,316
    Likes Received:
    3,819
    Trophy Points:
    431
    "Double performance per watt over Maxwell if you double the watts."

    [​IMG]
     
  10. tgipier

    tgipier Notebook Deity

    Reputations:
    197
    Messages:
    1,596
    Likes Received:
    1,572
    Trophy Points:
    181
    And triple the price ofcourse......
     
    TomJGX, HTWingNut and n=1 like this.
Loading...
Thread Status:
Not open for further replies.

Share This Page