Sign in with
Sign up | Sign in

Report: Nvidia Prepping Maxwell-based 750 Ti for February

By - Source: SweClockers | B 75 comments

Could we see our first Maxwell card in just a few weeks?

It seems like only last month, Nvidia's Jen-Hsun Huang was talking to us about the company's desktop GPU roadmap and Maxwell from the stage at GTC. Huang revealed that Maxwell would offer UVM, or unified virtual memory, which allows the CPU and GPU to share the same memory. If you can believe it, that was nearly a year ago. Since then, Nvidia has been keeping pretty quiet when it comes to information on Maxwell, but we always have the rumor mill to keep us going, right?

 

The latest scuttlebutt says we could see the first Maxwell-based GPU as early as next month. SweClockers says it will be a TSMC-made GeForce GTX 750 Ti manufactured on the 28 nanometer process, and it will replace the GTX 650 Ti Boost. The card is pegged for a February 18 launch, though Nvidia isn't confirming anything. You'll notice this card is carrying the GTX 700 series branding as opposed to the much talked about GTX 800 series branding that was rumored for Maxwell. TechPowerUp reckons Nvidia might be testing the waters with Maxwell on the existing 28 nanometer process before taking things to the next level on the future 20 nanometer nodes.

We've actually been hearing about Maxwell for almost four years. Way back in 2010, when Nvidia was shipping Fermi, the company named Kepler and then Maxwell, which promised 16 times the performance in parallel graphics-based computing and the ability to work independent of the CPU, as the next stops on its GPU roadmap. Looks like the wait is almost over!

Follow Jane McEntegart @JaneMcEntegart. Follow us @tomshardware, on Facebook and on Google+.

Discuss
Ask a Category Expert

Create a new thread in the News comments forum about this subject

Example: Notebook, Android, SSD hard drive

This thread is closed for comments
  • 2 Hide
    thesuperguy , January 18, 2014 8:33 PM
    I was admittedly hoping it would replace the 660.
  • 4 Hide
    amk09 , January 18, 2014 9:05 PM
    You know with this new architecture, it just might. If this 750 can hold it's own against the 660/560, and stays within the typical x50 price range, it could turn out to be one of the best bang for buck cards we've seen in this range of performance. Perfect entry level card for a console gamer wanting graphics that will match and exceed the quality of graphics that the new consoles generate.
  • -3 Hide
    mrcoolbreeze704 , January 18, 2014 9:16 PM
    But will it beat Hawaii based amd offerings?
  • Display all 75 comments.
  • 0 Hide
    dragonsqrrl , January 18, 2014 9:24 PM
    Quote:
    You know with this new architecture, it just might. If this 750 can hold it's own against the 660/560, and stays within the typical x50 price range, it could turn out to be one of the best bang for buck cards we've seen in this range of performance. Perfect entry level card for a console gamer wanting graphics that will match and exceed the quality of graphics that the new consoles generate.
    660 type performance would be my guess. Something to slot in right under the 760 in their lineup, probably targeting $150 - $200 range.
  • 4 Hide
    DarkSable , January 18, 2014 10:30 PM
    On the other hand, which gets me more excited... this means there's a STRONG chance we're going to be seeing full-fledged, 20nm Maxwell cards dropping soon-ish. :3
  • -1 Hide
    de5_Roy , January 19, 2014 12:11 AM
    x50TI (GM117), means 128-192bit memory bus max., 128bit more likely. i expect this one to be memory choked like 650ti boost. current kepler gpus could easily perform higher with 256bit bus, but nvidia cripples them so that they can gouge customers (gtx 660, 660ti).
    and "
    Quote:
    TechPowerUp reckons Nvidia might be testing the waters with Maxwell on the existing 28 nanometer process before taking things to the next level on the future 20 nanometer nodes.
    "? doesn't that mean qualcomm and apple fab-blocked nvidia? tsmc is already going into high volume production for 20nm arm socs ( seemingly from vendors that are not nvidia, lol).
  • 0 Hide
    hannibal , January 19, 2014 3:42 AM
    With the same production node they can not make this much quicker without making this chip bigger. 5-10% top more speed than 650. (if the size remains the same.) As someone said, this is a test run, and "real" 800 series comes later with smaller production node. What this allso means, is that 20nm production node is not quite in there just yeat. Interesting to see how long it will take untill it will be out and running in highend GPU production.
  • 0 Hide
    wdmfiber , January 19, 2014 5:13 AM
    TSCM was ramping up to 20nm last October:http://www.tomshardware.com/news/tsmc-20-nm-lithography-hkmg,24487.htmlI'm looking forward to the high ends cards(200 - 300 watts, 20nm), likely only a few months away.
  • 8 Hide
    Mousemonkey , January 19, 2014 5:18 AM
    Quote:
    But will it beat Hawaii based amd offerings?


    What? The mid range ones that don't exist? :heink: 
  • -7 Hide
    ElMoIsEviL , January 19, 2014 5:40 AM
    People still care about what nVIDIA has in the works? How quaint :) I would have thought that their lack of keeping up with the direction in which the industry is heading would turn most of its fans off. I mean proprietary closed ended solutions are just so 2004.Oh well... I guess they'll just be as surprised as the 3Dfx fans were when that company went belly up.
  • 7 Hide
    Mousemonkey , January 19, 2014 5:45 AM
    Quote:
    People still care about what nVIDIA has in the works? How quaint :) I would have thought that their lack of keeping up with the direction in which the industry is heading would turn most of its fans off. I mean proprietary closed ended solutions are just so 2004.Oh well... I guess they'll just be as surprised as the 3Dfx fans were when that company went belly up.


    Considering the implications of a current lawsuit and Nvidia's current bank balance I can't see them going belly up any time soon but then I'm also happy to leave it to history to record events.
  • -1 Hide
    FunSurfer , January 19, 2014 7:33 AM
    The most interesting thing about the 750 Ti is its power consumption, according to the graph above it should be a 1/3 than Kepler's. Also I hope Nvidia is working on a solution to the AC4 BF Physx performance (the GTX880Ti)
  • 0 Hide
    bison88 , January 19, 2014 9:49 AM
    Quote:
    The most interesting thing about the 750 Ti is its power consumption, according to the graph above it should be a 1/3 than Kepler's. Also I hope Nvidia is working on a solution to the AC4 BF Physx performance (the GTX880Ti)
    Quote:
    The most interesting thing about the 750 Ti is its power consumption, according to the graph above it should be a 1/3 than Kepler's. Also I hope Nvidia is working on a solution to the AC4 BF Physx performance (the GTX880Ti)
    Yeah and that most likely isn't going to be seen with the 750 Ti. Even if it has 660 or 660 Ti performance, those power envelopes are already as tiny as you'll get on 28nm. In a good way.
  • 0 Hide
    unksol , January 19, 2014 9:50 AM
    its a RUMOR and the big "news" of this rumor is shared GPU and CPU memory. which with nvidia must mean system memory. um.... AMD already did HSA and actually makes CPUs/GPUs...

    and their first card would be a midrange card no one with current gen would go for? sounds really week to me
  • 2 Hide
    Mousemonkey , January 19, 2014 9:59 AM
    Quote:
    its a RUMOR and the big "news" of this rumor is shared GPU and CPU memory. which with nvidia must mean system memory. um.... AMD already did HSA and actually makes CPUs/GPUs...

    and their first card would be a midrange card no one with current gen would go for? sounds really week to me


    I thought it was going to have an embedded Denver processor and if so then wouldn't that share the the same memory as the GPU and not the system? And which AMD card also has a CPU on it?
  • 0 Hide
    west7 , January 19, 2014 10:43 AM
    i wounder how much the performance gain will be? if 750ti have some power as 650ti/660 no one will buy them and i think it will cost lot more then kepler
  • 2 Hide
    Mousemonkey , January 19, 2014 10:49 AM
    Quote:
    i wounder how much the performance gain will be? if 750ti have some power as 650ti/660 no one will buy them and i think it will cost lot more then kepler

    I want to see how this :- " The next big feature is that Maxwell GPUs will embed a 64-bit ARM CPU core based on NVIDIA's "Project Denver." This CPU core will allow the GPU to reduce dependency on the system's main processor in certain GPGPU scenarios. Pole-vaulting the CPU's authority in certain scenarios could work to improve performance" is going to turn out.

    http://www.techpowerup.com/196956/nvidia-readies-geforce-gtx-750-ti-based-on-maxwell.html
  • -1 Hide
    redeemer , January 19, 2014 11:14 AM
    Quote:
    Quote:
    People still care about what nVIDIA has in the works? How quaint :) I would have thought that their lack of keeping up with the direction in which the industry is heading would turn most of its fans off. I mean proprietary closed ended solutions are just so 2004.Oh well... I guess they'll just be as surprised as the 3Dfx fans were when that company went belly up.


    Considering the implications of a current lawsuit and Nvidia's current bank balance I can't see them going belly up any time soon but then I'm also happy to leave it to history to record events.



    Hey Mouse, yeah Nvidia has a ton of Money however their avenues in the consumer market are dwindling. They wont succeed in mobile thanks to Qualcomm, Apple, Samsung. Another big blow was the loss on the professional side Mac Pros now use FirePro. Devs will now be optimizing for AMD

    Nvidia will retreat from the consumer market in the coming years. Competition is very fierce
  • -4 Hide
    redeemer , January 19, 2014 11:17 AM
    Quote:
    Quote:
    People still care about what nVIDIA has in the works? How quaint :) I would have thought that their lack of keeping up with the direction in which the industry is heading would turn most of its fans off. I mean proprietary closed ended solutions are just so 2004.Oh well... I guess they'll just be as surprised as the 3Dfx fans were when that company went belly up.


    Considering the implications of a current lawsuit and Nvidia's current bank balance I can't see them going belly up any time soon but then I'm also happy to leave it to history to record events.


    AMD Mullins
    4.5w Quad core X86
    3X CU 192 unit
    GCN

    K1
    5W not confirmed more like 10w
    Dual core ARM
    Kepler 192 units


    Qualcomm 805 is suppose to boost gprahics by over 40% over existing 800 series chip.

    As far as HSA goes Nvidia is years behind the competition
Display more comments