Home Reviews Articles News Forums Charts Techwiki Store
  • Latest News

    • XFX Retail GTX480 & GTX 470 Cards Pictured
    • Scythe Launches Yasya CPU Cooler
    • PowerColor Launches HD5770 Eyefinity 5 Card
    • LSI Starts Shipping LSISSS6200 PCI-E Solid State Storage Card
    • NVIDIA Slips in 197.13 BETA Drivers
    • AMD Phenom II X6 “Thuban” Clock Speeds Leaked
    • Southpaws Unite – Razer Develops Unique DeathAdder Gaming Mouse
    • ZOTAC Unleashes World’s First ION Upgrade Kits
  • Latest Reviews

    • XFX Radeon HD 5830 1GB GDDR5 Review
    • Gelid Tranquillo CPU Cooler Review
    • ASUS ROG Rampage III Extreme LGA1366 Motherboard Preview
    • ASUS Crosshair IV Formula AM3 890FX Motherboard Sneak Peak
    • Intel Core i7-980X Gulftown Six-Core 32nm Processor Review
    • ASUS HD 5870 Matrix 2GB & ASUS HD 5870 “V2″ Sneak Peek
    • NVIDIA GeForce GT 240 Roundup (EVGA, Gigabyte, MSI, Sparkle)
    • ASUS RoG Maximus III Extreme LGA 1156 Motherboard Preview
  • Featured Reviews

    • NVIDIA’s GeForce GF100 Under the Microscope
    • Cooler Master 690 II Advanced Case Review
    • Intel Westmere 32nm Launch & Clarkdale Core i5 661 CPU Review
    • 9-Way 850W Power Supply Roundup
    • Kingston SSDNow V Series 40GB SSD Review
    • EVGA P55 FTW LGA1156 Motherboard Review
    • Sapphire Radeon HD 5970 2GB OC Edition Review
    • XFX Radeon HD 5850 1GB GDDR5 Review

NVIDIA GT300 – Codename Fermi – Detailed  

Home > News > Video Card News > 
  Bookmark and Share
Posted by FiXT— September 30th, 2009, 10:47 AM

nvidia logo8 300x233 NVIDIA GT300   Codename Fermi   DetailedA few features have finally be put to the blank face that has only been known as the GT300. A few key details of the upcoming NVIDIA generation of video cards have been leaked, the first of which is a name. Codenamed Fermi, after the developer of the first nuclear power reactor, Enrico Fermi,  it appears to be quite suiting when looking at the rumoured specs.


Sources say that the card will be build on the 40nm manufacturing process and  feature 3 billion transistors, a 384-bit memory interface running on approximately 1.5GB  (3GB and 6GB versions also rumoured)  of GDDR5 memory. Taking GPU computing in a new direction, NVIDIA is redeveloping  ”shader cores” into CUDA cores, of which the card will feature 512. There are 32 of CUDA cores packed into each of the described “Shader Clusters” for a total of 16 clusters. The beauty of this new design is that these clusters are able to operate independently and balance load on calculations.


In another unique approach, they have added 1MB of L1 Cache onto the card, and divided it into 16KB chunks along with 786KB of unified L2 cache. NVIDIA is clearly pushing to embrace the video card as the new power house of the computer, and develop its parallel computing technology in mainstream video cards while still maintaining dominance in the visual graphics.


The icing on the cake is the native language and API support for CUDA (C), C++, Fortran,  DirectCompute 11, DirectX 11, OpenGL 3.1, and OpenCL.


NVIDIA’s GTC conference is being held today, and we look forward to updating you later in the day as we hear more. It is rumoured that NVIDIA CEO Jensen Huang will be debuting a working model of the GT300 during his keynote.


For more technical details on the GT300 visit the Bright Side of News




Tags: 40nm, fermi, gf100, gt300, nvidia

Related News

  • NVIDIA Fermi GT300 Video Cards A Fake?
  • NVIDIA's Next-Gen Parts Get Named: GTX 480 & GTX 470
  • NVIDIA Officially Unveils Next Generation CUDA GPU Architecture - Codenamed "Fermi"
  • ASUS Developing a Beast - Dual GTX 285 4GB GPU
  • NVIDIA Launches 5 New Notebook GPU's; Takes 40nm and DX10.1 Mobile
 

Related Articles

Related Articles

  • NVIDIA Fermi GT300 Video Cards A Fake?
  • NVIDIA's Next-Gen Parts Get Named: GTX 480 & GTX 470
  • NVIDIA Officially Unveils Next Generation CUDA GPU Architecture - Codenamed "Fermi"
  • ASUS Developing a Beast - Dual GTX 285 4GB GPU
  • NVIDIA Launches 5 New Notebook GPU's; Takes 40nm and DX10.1 Mobile

Comments (21)

  1. Jake_HT's Avatar
    Jake_HT

    September 30, 2009 10:49 AM

    Sounds ridiculous...

  2. Phobia's Avatar
    Phobia

    September 30, 2009 10:54 AM

    3 billion transistors? thats ridiculous..

  3. BobbyLou's Avatar
    BobbyLou

    September 30, 2009 10:54 AM

    sad, and we alll thought ATI was going to pull an AMD again. :C

  4. martin_metal_88's Avatar
    martin_metal_88

    September 30, 2009 11:28 AM

    3 billionand guess what, she can drive your kids out when always giving a awesome performance of gamin in DX11 game. I wont belive before I see! maybe are they all 2 GPU card so 1.5+1.5 make 3?

    Ask for more.

  5. CMetaphor's Avatar
    CMetaphor

    September 30, 2009 11:32 AM

    IMHO, im not sure going for such insane performance is really beneficiary at this point. I mean, unless Nvidia comes out with a counter for the Eyefinity technology which allows far more pixels than a single 30'' to be driven at once, there is almost no need for such power. Think about it for a second: Why do you need 3 Billion transistors at 40nm with 1.5Gb of GDDR5... and you're only running a 24" panel @ 1900x1200? It doesn't make sense, this graphics card would be insane overkill. Now if Nvidia DOES come out with a counter, then it makes more sense. As soon as you start extra more monitors you need bigger boosts in graphics horsepower. We'll see just what Nvidia does here, perhaps there's a big magic white rabbit hidden somewhere in their top hat, hehe.

  6. Jake_HT's Avatar
    Jake_HT

    September 30, 2009 11:34 AM

    Quote:
    Originally Posted by CMetaphor View Post
    IMHO, im not sure going for such insane performance is really beneficiary at this point. I mean, unless Nvidia comes out with a counter for the Eyefinity technology which allows far more pixels than a single 30'' to be driven at once, there is almost no need for such power. Think about it for a second: Why do you need 3 Billion transistors at 40nm with 1.5Gb of GDDR5... and you're only running a 24" panel @ 1900x1200? It doesn't make sense, this graphics card would be insane overkill. Now if Nvidia DOES come out with a counter, then it makes more sense. As soon as you start extra more monitors you need bigger boosts in graphics horsepower. We'll see just what Nvidia does here, perhaps there's a big magic white rabbit hidden somewhere in their top hat, hehe.
    THink about the folding power!!!!

  7. Sushi Warrior's Avatar
    Sushi Warrior

    September 30, 2009 11:38 AM

    Eyefinity isn't really much, just glorified multi-monitor. You can use SLI multi-monitor, right? So that's up to 9 displays or so (3 plugs x3 cards?)

  8. CMetaphor's Avatar
    CMetaphor

    September 30, 2009 11:38 AM

    Quote:
    Originally Posted by Jake_HT View Post
    THink about the folding power!!!!
    Okay... so theyre going to create a 400-500$ (minimum) graphics card for folding only? thats just silly. There has to be a reason to strive for so much more power, beyond just the DX11 power requirement. Either the details we've gotten so far are bogus, or they have someting else up their sleeve to take full advantage of it.

    come to think of it, this card is probably going to be Extremely expensive...
    Unless I'm mistaken, its their first 40nm/GDDR5 graphics card ever, something that AMD has had for many months now (since the 4770). Going to be one hell of a premium to buy "3-billion transistors". hehe

  9. Perineum's Avatar
    Perineum

    September 30, 2009 11:44 AM

    Well it certainly sounds good....

    This is pretty much my first time reading prerelease stuff like this so I'm interested to see how much of this is FUD

  10. FiXT's Avatar
    FiXT

    September 30, 2009 11:50 AM

    Keep in mind, to use Eyefinity - you HAVE to have a DisplayPort monitor attached and you CAN'T use an adapter.

    I haven't managed to find an instock DP monitor yet, let alone one that isn't ridiculously overpriced compared to a regular input. + You have the chore of finding a cable which runs a good $30+



    I can't remember where I read it, but I feel as though this sums it up between the new ATI and NVIDIA cards:

    ATI's cards are an EVOLUTION. They have adapted their existing technology to enhance the gaming experience and visuals. NVIDIA is a REVOLUTION. They have overthrown the original way of thinking and the video card, and are seemingly creating a whole new card focusing on parallel computing and GPU based processing along with improved gaming and graphics. NVIDIA envisions, that the graphics card, can be the new major system component capable of running everything.

    It has potential to be a wild success.... or an utter failure.

  11. SKYMTL's Avatar
    SKYMTL

    September 30, 2009 12:20 PM

    The issue at this point is DisplayPort. Nothing more and nothing less. People have no issue going to the computer recycler or Ebay and picking up a trio of $100 22" monitors. Talk about a DP monitor though and you get a massive price increase. I am sure that the overall price will go down in the coming months but at this point, there is very little reason out there to recommend DisplayPort.

  12. Squeetard's Avatar
    Squeetard

    September 30, 2009 12:43 PM

    Quote:
    Originally Posted by CMetaphor View Post
    Okay... so theyre going to create a 400-500$ (minimum) graphics card for folding only? thats just silly. There has to be a reason to strive for so much more power, beyond just the DX11 power requirement. Either the details we've gotten so far are bogus, or they have someting else up their sleeve to take full advantage of it.
    You read the article right?

    Quote:
    The icing on the cake is the language and API native support for CUDA (C), C++, Fortran, DirectCompute 11, DirectX 11, OpenGL 3.1, and OpenCL.
    Quote:
    NVIDIA is clearly pushing to embrace the video card as the new power house of the computer,

  13. Sushi Warrior's Avatar
    Sushi Warrior

    September 30, 2009 12:46 PM

    The most intriguing thing to me is the L1 and L2 cache integrated into it....

  14. bojangles's Avatar
    bojangles

    September 30, 2009 12:49 PM

    Quote:
    Originally Posted by Sushi Warrior View Post
    The most intriguing thing to me is the L1 and L2 cache integrated into it....
    Which I think is a great idea. SIMD. The closer the data is (aka super fast memory) the faster the GPU can crunch.

  15. rjbarker's Avatar
    rjbarker

    September 30, 2009 01:12 PM

    Very much looking forward to this..!
    GTX 380 SLI ....perhaps Jan (after the Christams rush)...3 x GTX 280's For sale ;)

  16. Sagath's Avatar
    Sagath

    September 30, 2009 02:02 PM

    Quote:
    NVIDIA is clearly pushing to embrace the video card as the new power house of the computer,
    Power House? Not to mention "Power Drain, Space Heater" and my personal favorite "Consumes 4x as much wattage as the rest of your PC combined."

    Intel realized several years ago that Monolithic wasnt working. Why havnt ATi/nV?

  17. LionRed's Avatar
    LionRed

    September 30, 2009 04:01 PM

    Quote:
    Originally Posted by rjbarker View Post
    Very much looking forward to this..!
    GTX 380 SLI ....perhaps Jan (after the Christams rush)...3 x GTX 280's For sale ;)
    ^
    |
    |
    What he said... 2 gtx 295's for sale soon.

  18. muse108dc's Avatar
    muse108dc

    September 30, 2009 04:11 PM

    if ati's cards are close to these in performance with the significantly lower number of transistors I'd say that ATI had a nice advantage over nVidia

  19. fefox's Avatar
    fefox

    September 30, 2009 04:19 PM

    Quote:
    Originally Posted by FiXT View Post
    Codenamed Fermi, after the developer of the first nuclear power reactor,Enrico Fermi
    They forgot to mention it was his design of the cooling tower that will now be in your back yard that keeps the card cool

  20. MpG's Avatar
    MpG

    September 30, 2009 04:52 PM

    Quote:
    Originally Posted by CMetaphor View Post
    IMHO, im not sure going for such insane performance is really beneficiary at this point. I mean, unless Nvidia comes out with a counter for the Eyefinity technology which allows far more pixels than a single 30'' to be driven at once, there is almost no need for such power. Think about it for a second: Why do you need 3 Billion transistors at 40nm with 1.5Gb of GDDR5... and you're only running a 24" panel @ 1900x1200? It doesn't make sense, this graphics card would be insane overkill. Now if Nvidia DOES come out with a counter, then it makes more sense. As soon as you start extra more monitors you need bigger boosts in graphics horsepower. We'll see just what Nvidia does here, perhaps there's a big magic white rabbit hidden somewhere in their top hat, hehe.
    If you check out the Anand article regarding the specifics of Fermi, there's some seriously interesting stuff there, outside the realm of gaming. Realistically, I think I agree - our GPU's are outpacing our display needs, even among many/most enthusiasts. But it sounds like Nvidia is aggressively pursuing non-gaming applications, and there's definitely some potential there, if Nvidia can sell it to the customers who are looking for it. That aside, there seems to be some hinting that they could scale down this core a little easier, although who can tell with all the marketingese it's buried in?

  21. belgolas's Avatar
    belgolas

    September 30, 2009 05:00 PM

    Here is a cool video.

    YouTube - NVIDIA Fermi Architecture Interview - PC Perspective

(21) comments | Add your comments

All About Us
Contact Us
Terms of Use, Privacy
Awards
Help
Community
Forum
Links
YouTube Channel
Syndications
Reviews RSS Feeds
News RSS Feeds
Copyright © 2006-2009, HardwareCanucks.com, HWC, a Division of GTO Media Inc. All Rights Reserved. Terms & Conditions