New CELL Article Online

Article: CELL Microprocessor III
By: David Wang (, August 5, 2005 11:00 pm
Room: Moderated Discussions
David Wang ( on 8/3/05 wrote:
>Panajev2001a ( on 8/3/05 wrote:
>>David Wang ( on 8/2/05 wrote:
>>>fastpathguru ( on 8/2/05 wrote:
>>>>Re: The Mystery of the Missing FLOPS, is it possible that it's solved by accounting
>>>>for the extra meat in the PPE? I didn't see that mentioned as a possibility.
>>>You need 68 flops per cycle to get 218 GFlops @ 3.2 GHz.
>>>It's easy to get 56 of those flops from 7 SPE's, using FMADD, but then you'd need
>>>12 flops per cycle out of the PPE to get to 68.
>>>Again, assume that you can crank 8 Flops per cycle out with FMADD, you still need
>>>4 flops per cycle. The PPE is multi-threaded, but the impression that I got from
>>>the IBM folks is that the scalar and vector FP units are actually shared so you
>>>really can't get more flops out of the PPE either. Regardless, even if the FP units
>>>aren't shared, you still can't get 4 flops per cycle out of the scalar side of thing.
>>>So no, the beefier PPE can't get to 12 flops per cycle even if it had independent scalar/vector FP units.
>>Why not ? See Gekko ;).
>I've forgotten about Gekko entirely.
>So the 12 Flops/cycle can be attained by the PPE if the PPE's scalar FPU can do
>a 2 wide SP FMADD in addition to a separate and independent Altivec unit. That
>does make things quite curious. I'm not sure how many programmers will be able
>to make use of this capability, but I think you're right, the Gekko ISA extension
>to the DD2 PPE explanation sounds to be the most reasonable explanation that I have
>come across about how Sony was able to claim the 218 GFlops rating with only 7 SPE's @ 3.2 GHz.
>Perhaps you can write a rebuttal article and we can link it to the current article.
>I think the readership in general will benefit from this discussion.

I've since received two negative feedbacks on the "DD2 PPE has Gekko" speculation from people who *should* know. So I'm now back to square one. The flops are mysterious once again with no explanation as to how to account for the last 4 flops per cycle.

< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
New CELL Article OnlineDavid Kanter2005/08/02 10:32 AM
  New CELL Article Onlinemas2005/08/02 11:46 AM
    New CELL Article Onlinemas2005/08/02 11:53 AM
    New CELL Article OnlineDavid Wang2005/08/02 12:46 PM
      New CELL Article Onlinefastpathguru2005/08/02 03:05 PM
        New CELL Article OnlineDavid Wang2005/08/02 05:27 PM
          New CELL Article OnlinePanajev2001a2005/08/03 02:26 AM
            New CELL Article OnlineDavid Wang2005/08/03 10:28 AM
              New CELL Article OnlineDeadmeat2005/08/04 12:05 PM
                New CELL Article OnlineDavid Wang2005/08/04 04:47 PM
                  New CELL Article OnlineDeadmeat2005/08/04 06:04 PM
                    New CELL Article Onlinejohn evans2005/08/04 07:30 PM
                      New CELL Article OnlineDeadmeat2005/08/05 11:10 AM
                        New CELL Article OnlineLinus Torvalds2005/08/05 05:21 PM
                          New CELL Article OnlineDeadmeat2005/08/05 06:33 PM
                            New CELL Article Onlinefastpathguru2005/08/05 09:36 PM
                              New CELL Article Onlinejohn evans2005/08/05 09:51 PM
                              New CELL Article OnlineDeadmeat2005/08/06 03:09 AM
                                New CELL Article Onlinefastpathguru2005/08/06 05:29 AM
                                  New CELL Article OnlineDeadmeat2005/08/07 03:06 PM
                    New CELL Article OnlineDavid Wang2005/08/04 08:03 PM
                      New CELL Article OnlineDeadmeat2005/08/05 11:21 AM
                        New CELL Article OnlineDavid Wang2005/08/05 10:51 PM
              New CELL Article OnlineDavid Wang2005/08/05 11:00 PM
                New CELL Article OnlineDeadmeat2005/08/07 02:39 PM
                  New CELL Article OnlineDavid Wang2005/08/08 12:57 PM
                    New CELL Article OnlineDeadmeat2005/08/08 01:55 PM
                      New CELL Article OnlineDavid Wang2005/08/08 02:37 PM
                        New CELL Article OnlineDeadmeat2005/08/08 04:05 PM
                          New CELL Article OnlineDavid Wang2005/08/08 04:47 PM
                            New CELL Article OnlineDeadmeat2005/08/08 05:25 PM
                              Implausible at best, irrational most likely...David Kanter2005/08/08 05:51 PM
                                Implausible at best, irrational most likely...Deadmeat2005/08/09 09:26 AM
                              New CELL Article OnlineDavid Wang2005/08/08 06:46 PM
                                New CELL Article OnlineDeadmeat2005/08/09 09:36 AM
                                  New CELL Article OnlineDavid Wang2005/08/09 10:12 AM
                                    New CELL Article OnlineDeadmeat2005/08/09 12:26 PM
                                      New CELL Article OnlineDavid Wang2005/08/09 01:36 PM
                                New CELL Article OnlineAaron Spink2005/08/09 01:57 PM
                                  New CELL Article OnlineDavid Wang2005/08/10 09:06 AM
                    New CELL Article OnlineSerge Monkewitz2005/08/09 12:18 PM
                      New CELL Article OnlineDeadmeat2005/08/09 12:30 PM
                        New CELL Article OnlineVitaly Vidmirov2005/08/11 12:36 AM
      New CELL Article OnlineAnonymous2005/08/03 03:11 PM
        New CELL Article Onlinefastpathguru2005/08/03 04:19 PM
          New CELL Article Onlinemas2005/08/03 06:59 PM
            New CELL Article OnlineJosé Javier Zarate2005/08/04 04:20 AM
              New CELL Article Onlinemas2005/08/04 04:27 AM
          New CELL Article Onlinemas2005/08/05 05:50 AM
  New CELL Article OnlinePiedPiper2005/08/02 08:02 PM
Reply to this Topic
Body: No Text
How do you spell avocado?