Page 2 of 10 FirstFirst 1234 ... LastLast
Results 11 to 20 of 91

Thread: GCN FireStream ... still not available?

  1. #11
    Senior Member
    Join Date
    Jul 2011
    Posts
    2,056
    W9000 = Tahiti @1GHz?

    AMD clearly realises they underclocked their cards if they are going to release a professional card with higher clocks.

  2. #12
    Yes, clear indication that 7970 GHz will be clocked at 1.1 or so.

  3. #13
    640k who needs more?
    Join Date
    Apr 2011
    Posts
    762
    So we finally have confirmation Tahiti is hardware capable of 1/4 D.P. rate, rather than the 1/2 D.P. some had supposed.

  4. #14
    8-bit overflow
    Join Date
    Apr 2012
    Posts
    483
    Quote Originally Posted by kalelovil View Post
    So we finally have confirmation Tahiti is hardware capable of 1/4 D.P. rate, rather than the 1/2 D.P. some had supposed.
    its pretty obvious they weren't going for 1/2 DP. Even nvidia is abandoning 1/2 DP since its much too resource heavy.

  5. #15
    640k who needs more?
    Join Date
    Apr 2011
    Posts
    762
    Quote Originally Posted by esrever View Post
    its pretty obvious they weren't going for 1/2 DP. Even nvidia is abandoning 1/2 DP since its much too resource heavy.
    AMd has mentioned however that the GCN architecture is theoretically capable of 1/2 DP, the capability to do so just isn't in any current chips.

  6. #16
    8-bit overflow
    Join Date
    Apr 2012
    Posts
    483
    Where did they say that? it would be weird to have the feature but not implement it.

  7. #17
    640k who needs more?
    Join Date
    Apr 2011
    Posts
    762
    Quote Originally Posted by esrever View Post
    Where did they say that? it would be weird to have the feature but not implement it.
    "Elsewhere we?ve already mentioned FP64 support. All GCN GPUs will support FP64 in some form, making FP64 support a standard feature across the entire lineup. The actual FP64 performance is configurable ? the architecture supports ? rate FP64, but ? rate and 1/16 rate are also options. We expect AMD to take a page from NVIDIA here and configure lower-end consumer parts to use the slower rates since FP64 is not currently important for consumer uses"

    http://www.anandtech.com/show/4455/a...-for-compute/6

  8. #18
    8-bit overflow
    Join Date
    Apr 2012
    Posts
    483
    I don't see why they would keep that from their top of the line firepros, weird. If AMD has 1/2 DP then there would be no way kepler can compete in compute as the gk110 will not be able to do 2TFLOP DP.

  9. #19
    Quote Originally Posted by esrever View Post
    its pretty obvious they weren't going for 1/2 DP. Even nvidia is abandoning 1/2 DP since its much too resource heavy.
    If you take 2 arrays of 32 CUDA cores away from the SMX and say goodbye to 2-way co-issue of ALU instructions, you will get 1/2 DP rate.
    Anyway, the problem could be more likely register file to execution units wiring complexity I think?

  10. #20
    8-bit overflow
    Join Date
    Apr 2011
    Posts
    324
    w9000 dual gpu? (see back of card)
    6gb , 4 TFLOPs single , 1 TFLOPs dual

    http://www.4gamer.net/games/135/G013536/20120615004/

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
WordPress Appliance - Powered by TurnKey Linux