1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Rumor: Entire NVIDIA Kepler 600-Series Specs, Price And Release Dates

Discussion in 'NVIDIA' started by qubit, Feb 7, 2012.

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

    qubit Overclocked quantum bit

    Joined:
    Dec 6, 2007
    Messages:
    14,390 (4.06/day)
    Thanks Received:
    8,084
    Location:
    Quantum Well UK
    For those that would like to read some juicy leaked info on NVIDIA's next gen and want it with the qubit treatment, have a look at the news article I've just published at <link removed>.

    From what I can tell by searching, there isn't a TPU version of this article, so this should be ok to put here.

    EDIT: I'm sorry people, it was against forum rules promoting my own article, no matter how well intentioned. :eek: It was my mistake and I've been corrected by a couple of mods.
     
    Last edited: Feb 7, 2012
  2. LAN_deRf_HA

    LAN_deRf_HA

    Joined:
    Apr 4, 2008
    Messages:
    4,642 (1.36/day)
    Thanks Received:
    1,001
    If that's true that vram stinks. I was pushing the limit on my 3GB 580 day one. So that means you have to hold out for the 4 GB version, which will be like $700 bucks and come months later. And I really would hold out. If you're spending that much you should be able to run anything, mods and whatever else.
     
  3. entropy13

    entropy13

    Joined:
    Mar 2, 2009
    Messages:
    5,026 (1.63/day)
    Thanks Received:
    1,273
    2GB's still not much to scoff at. Law of Diminishing Returns and all that. There's quite an improvement between 512MB and 1GB, some improvement in some resolutions between 1GB and 2GB, and HD 7970's 3GB isn't a factor at all unless you go multi-monitor.
     
  4. m1dg3t

    m1dg3t

    Joined:
    May 22, 2010
    Messages:
    2,516 (0.95/day)
    Thanks Received:
    708
    Location:
    Canada
    Look's like the same "Lenzfire" graph's to me.

    You can polish a turd all you want but it's still a turd :shadedshu
     
  5. ViperXTR

    ViperXTR

    Joined:
    Jan 31, 2011
    Messages:
    1,690 (0.71/day)
    Thanks Received:
    543
    yeah, posted days ago (on the kepler thread) shader clocks? i thought they were a goner?
     
  6. Funtoss

    Funtoss New Member

    Joined:
    Mar 4, 2011
    Messages:
    588 (0.25/day)
    Thanks Received:
    61
    Location:
    New Zealand
    i wanna see the 660 in action!! oh and also those high end 600 series as well ;)
     
  7. borden5 New Member

    Joined:
    Apr 4, 2011
    Messages:
    66 (0.03/day)
    Thanks Received:
    11
  8. Steevo

    Steevo

    Joined:
    Nov 4, 2005
    Messages:
    9,812 (2.28/day)
    Thanks Received:
    2,265
    15% fewer shaders at the same speed perform 25% worse, and 50% less shaders perform the same as a 580, which is 10-15% slower than the 7970 at resolutions we have used since 2000, at a higher speed.

    And it only has 4 million transistors VS 6 million? die size to be 50% larger for 45% more performance?

    They could make the world stop by just using 60% fewer shaders, with their scaling on @ 850Mhz it would have 0% performance!!!!
     
    10 Year Member at TPU 10 Million points folded for TPU
  9. Crap Daddy

    Crap Daddy

    Joined:
    Oct 29, 2010
    Messages:
    2,940 (1.18/day)
    Thanks Received:
    1,362
    Just rumors although I'm surprised so many sites have taken lenzfire's "leak" seriously.
     
  10. damric

    damric

    Joined:
    Feb 17, 2010
    Messages:
    503 (0.18/day)
    Thanks Received:
    152
    I only buy rumors from the trusted expert rumor trolls at dominamhabererer.
     
  11. W1zzard

    W1zzard Administrator Staff Member

    Joined:
    May 14, 2004
    Messages:
    16,768 (3.46/day)
    Thanks Received:
    17,507
    i think bta looked into the leaked info and found it to be not reliable
     
    10 Year Member at TPU
  12. the54thvoid

    the54thvoid

    Joined:
    Dec 14, 2009
    Messages:
    6,397 (2.28/day)
    Thanks Received:
    5,654
    Location:
    Glasgow - home of formal profanity
    Especially given:

    a) the graph looks tawdry and very amateur (colours, fonts, everything).
    b) the TBD (to be decided/declared) usage. So it's a case of fakes being made by cobbling together every other Kepler rumour out there.

    Meh.

    There is no Kepler news.
     
  13. Maban

    Maban

    Joined:
    Mar 6, 2008
    Messages:
    2,691 (0.78/day)
    Thanks Received:
    1,363
    Location:
    Minnesota
    The lack of GK107, a chip known to be in the works, is suspicious.
     
  14. radrok

    radrok

    Joined:
    Oct 26, 2011
    Messages:
    3,115 (1.47/day)
    Thanks Received:
    890
    We'll probably see a custom version of the Nvidia top dog with 4GB Vram, just like 1,5GB -> 3GB on the GTX 580.
     
  15. EarthDog

    EarthDog

    Joined:
    Dec 31, 2009
    Messages:
    9,863 (3.54/day)
    Thanks Received:
    5,206
    Location:
    Ohio
    Interesting take...on (likely false) leaked info...

    I dont see the reason to be disappointed in 2GB of VRAM considering that even BF3 at 1920x1200 full tilt only uses 1.2GB (at least thats what AB shows me when playing Ultra, HBAO, 4xAA/16xAF). Certainly though, with a multi monitor setup this may actually be a concern, but that point wasnt mentioned. 2GB single monitor is plenty it seems. If this card can push 5760x1200 at palyable frame rates, it could very well be a concern.

    Whats is this 'power of 2' stuff and what are you basing this claim of inefficiency when you are not using it come from? I understand binary/base2 and all...but I do not know the in's and out's of it or why things being the power of 2 would make it more efficient. I have never heard (may be out there of course, the webz r big!) from any reviewer, say that before (as in, 'this card would be better/'max out' if it was using the power of 2'). Can you please link up where you got that information from so I may learn? Thanks!

    EDIT: Are there any cards in this leaked info that do not match the 'power of 2' and would therefore be more inefficient?
     
    Last edited: Feb 7, 2012
  16. Red_Machine

    Red_Machine

    Joined:
    Oct 1, 2010
    Messages:
    2,125 (0.85/day)
    Thanks Received:
    589
    Location:
    Marlow, ENGLAND
    If this is true, then I may have to replace my 580.

    Double the shaders, man...
     
  17. phanbuey

    phanbuey

    Joined:
    Nov 13, 2007
    Messages:
    5,895 (1.65/day)
    Thanks Received:
    1,426
    Location:
    Austin Texas
    yeap lets see... April is still a ways off though, so it will be fun to see pricing. Also Ivy Bridge is coming out then, so I might hold off even longer to upgrade VC.
     
  18. btarunr

    btarunr Editor & Senior Moderator Staff Member

    Joined:
    Oct 9, 2007
    Messages:
    33,783 (9.38/day)
    Thanks Received:
    17,263
    Location:
    Hyderabad, India
    There's a reason we didn't cover that Lenzfire leak...it's fake. Next time, submit news, and we'll take a look at it. No promoting other sites.
     
    qubit says thanks.
  19. qubit

    qubit Overclocked quantum bit

    Joined:
    Dec 6, 2007
    Messages:
    14,390 (4.06/day)
    Thanks Received:
    8,084
    Location:
    Quantum Well UK
    I'm sorry about that and I've added an apology to my OP. :)

    Sure about Lenzfire, I'm not surprised, that's why I said about keeping the skeptical hat on! (especially when I saw it wasn't posted here. ;) )
     

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)

Thread Status:
Not open for further replies.