Results 1 to 3 of 3

Thread: Nvidia 40nm Desktop GPUs Line-Up For 2009

  1. #1
    Join Date
    Oct 2008
    Posts
    421

    Nvidia 40nm Desktop GPUs Line-Up For 2009

    As foreshadowed the announcement in November of TSMC foundry, which indicated that it began mass production of 40 nanometer chips, NVIDIA prepare a full line of chips using that manufacturing process for the year 2009, d According to the online magazine anglophone VR-Zone.

    The majority of graphics chips from ATI (a division of AMD) and NVIDIA are currently reminder engraved in 55 nm and 65 nm. The reduction in the fine engraving helps reduce production costs, energy consumption and size of chips. A size, a bullet engraved in 40 nm is more powerful than other 65-nm.

    The chip stamped GT212 and succeed in the second quarter of 2009 with different versions of the GT200 cards qu'intègrent GeForce GTX 260 and 280 and the future GTX 285 and 295, which likely will have a life rather short.


    It was not until the third quarter to see off the 40 nm segments of entry and mid-range. The GT214 is expected to replace the G94 used by the GeForce 9600 GT, the GT216 will take on him instead of the G96 (GeForce 9500 GT), while the GT218 supplant the G98 (GeForce 9400 GT). The iGT209 for its part should take the relay chip integrated chipsets for GeForce 9300 and 9400.

    The GT300 series supporting DirectX 11 could finally see the day from the fourth quarter of 2009 on the segment of the premium.

  2. #2
    Join Date
    Oct 2008
    Posts
    56

    Re: Nvidia 40nm Desktop GPUs Line-Up For 2009

    Given that ATI plans to bring 40nm in Q1 or early Q2, this schedule looks like they might be a bit behind. ATI also plans to bring 40nm to mobile early in 2009 as well, so there's another area that is rather quiet on the Nvidia front. Have to see how '09 pans out.

  3. #3
    Join Date
    Oct 2008
    Posts
    421

    Nvidia 40nm chips to be similar to GT200

    Next generation Nvidia in the 40nm manufacturing process will share the basic design concept with the GT200 design. This architecture was introduced at 65nm and was rather hot, and forced Nvidia to make very big chips, but eventually the architecture was flexible enough to shrink to 55nm and bring some new products such as GTX 260, 285 and 295 all based on 55nm GT200b chips. The original idea was carried from 65nm to 55nm.

    The future will bring a similar concept but at 40nm, and you can easily expect more shaders, higher clocks and less heat dissipation from this new GPUs. Knowing Nvidia there should be something innovative, but many sources have confirmed that the 40nm design basics will be close to GT200 concept. Nvidia bets a lot in 2009 on Cuda and PhysX and some innovations are expected in this field. Nvidia will probably try to optimise its GPUs further for parallel computing.

    Nvidia will still stick to big monolithic chips for the high end, pushing the transistor count and hopefully performance limits to the next level, and it always has a safe net woven of two GPUs on a single card. If ATI’s new generation 40nm dual GPU card, lets call it R800 ends up faster than single 40nm high end Nvidia chip, Nvidia will use two of its chips for its top range card, and the game is on.

Similar Threads

  1. NVIDIA GeForce GTX 700M series GPUs Announced
    By Damiano in forum Web News & Trends
    Replies: 1
    Last Post: 03-06-2013, 10:28 AM
  2. Smallest GPUs for NVIDIA and ATI coming soon
    By RichiRich in forum Monitor & Video Cards
    Replies: 3
    Last Post: 21-02-2009, 10:43 PM
  3. N10x, the symbol of NVIDIA mobile GPUs
    By Eaton in forum Portable Devices
    Replies: 2
    Last Post: 07-01-2009, 03:42 PM
  4. NVIDIA Launches the GeForce 9 Series Motherboard GPUs
    By Reegan in forum Motherboard Processor & RAM
    Replies: 1
    Last Post: 16-10-2008, 07:09 PM

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Page generated in 1,713,585,043.01755 seconds with 17 queries