90nm A64 Review

By: IntelUser2000 (Intel_user2000.delete@this.yahoo.ca),
Room: Moderated Discussions
Peter Gerassimoff (PGERASSIMOFF@wi.rr.com) on 9/25/04 wrote:
---------------------------
>Paul DeMone (pdemone@igs.net) on 9/24/04 wrote:
>---------------------------
>>Groo (charlie@theinquirer.net) on 9/24/04 wrote:
>>---------------------------
>>>How about Dothan ramping?
>>
>>How about it?
>>
>>Banias has a 25 W TDP at top speed of 1.7 GHz
>>Dothan was intro'd with a 21 W TDP at 2.0 GHz
>>
>>Athlon 64 has a 89 W TDP at top speed of 2.4 GHz
>>in 130 nm.
>>
>>Unless AMD intro's 90 nm A64 with 2MB L2 at under
>>200 mm2 that runs 75 W TDP at 2.8 GHz then one
>>has grounds to say that relative to Intel's 90 nm,
>>AMD's 90 nm process is a disappointment. The
>>unkind might even say "broken". ;-)
>
>Broken, HA! 90nm A64 3000+ 939 1.8GHz at 1.2V uses 16-20W, less than Dothan's
>21.5W TDPtyp. 90nm A64 3500+ 939 2.2GHz at 1.4V uses 25-32W. Dothan left well behind in performance.
>
>See "Some 90nm A64 Power Data".
>
>Pete
>

The fact that Dothan runs at 1.7Ghz with no fan indicates that Dothan runs with much less TDP than 90nm A64 since even with only 87mm2 die, it runs cooler than A64 with over 100mm2 die.

Link: http://www.x86-secret.com/articles/cm/g5m100-n/itxpm-9.htm
< Previous Post in ThreadNext Post in Thread >
Thread (106 posts)
TopicPosted ByPosted
90nm A64 ReviewBeelzebubba9
  90nm A64 Reviewmas
    90nm A64 ReviewBeelzebubba9
      90nm A64 another 'review'mas
        90nm A64 another 'review' o/cmas
  90nm A64 ReviewGroo
    90nm A64 ReviewDean Kent
    90nm A64 ReviewBeelzebubba9
      90nm A64 Reviewmas
        90nm A64 ReviewMarc M.
          90nm A64 Reviewmas
            90nm A64 ReviewMarc M.
              90nm A64 ReviewBill Todd
                90nm A64 ReviewDavid Wang
                  90nm A64 ReviewPeter Huisken
                    90nm A64 ReviewPaul DeMone
                      90nm A64 ReviewChuck
                        90nm A64 ReviewMarc M.
                        90nm A64 ReviewPaul DeMone
                          90nm A64 ReviewSingh, S.R.
                            90nm A64 ReviewDavid Wang
                              90nm A64 ReviewSingh, S.R.
                                90nm A64 ReviewPaul DeMone
                                  90nm A64 ReviewSingh, S.R.
                                    90nm A64 Reviewqzm
                                  Counting transistors in the coreDoug Siebert
                                    Counting transistors in the corePaul DeMone
                                    Counting transistors in the corePeter Huisken
                      90nm A64 ReviewDavid Wang
                        90nm A64 ReviewDoug Siebert
                          90nm A64 ReviewJouni Osmala
                        90nm A64 ReviewPaul DeMone
                          Scaling to the barrier of SiliconJohn H
                            Scaling to the barrier of SiliconPaul DeMone
                          90nm A64 ReviewDavid Wang
                          90nm A64 ReviewMartin Høyer Kristiansen
                            90nm A64 Reviewjohn evans
                            90nm A64 ReviewPaul DeMone
                              90nm A64 ReviewMartin H. Kristiansen
                              90nm A64 ReviewChuck
                                90nm A64 ReviewPaul DeMone
                                  90nm A64 ReviewMatt Craighead
                                    Actually the reverse seems to be trueIlleglWpns
                                    Exactly! Load+update is suboptimal 4 jagged types. (NT)Chuck
                                    ia64 scheduling.. it is the point.qzm
                                      ia64 scheduling.. it is the point.Chuck
                                        ia64 scheduling.. it is the point.qzm
                                        ia64 scheduling.. it is the point.Matt Craighead
                                          ia64 scheduling.. it is the point.qzm
                                            ia64 scheduling.. it is the point.Anonymous
                                              ia64 scheduling.. it is the point.Matt Craighead
                                                ia64 scheduling.. it is the point.Paul DeMone
                                                  ia64 scheduling.. it is the point.john evans
                                                  ia64 scheduling.. it is the point.Chuck
                                                  ia64 scheduling.. it is the point.Anonymous
                                            ia64 scheduling.. it is the point.Matt Craighead
                                              General question on this issuePaul DeMone
                                                General question on this issuejohn evans
                                                General question on this issueMichael S
                                                sum addressing not known in 90s when ia64 was madeChuck
                                                  Sum addressing not a substitutePaul DeMone
                                                    Yes it isChuck
                                                      [O.T.] Non-speculative L1D accessMichael S
                                                        [O.T.] Non-speculative L1D accessChuck
                                                          [O.T.] Non-speculative L1D accessMatt Craighead
                                                            [O.T.] Non-speculative L1D accessMichael S
                                                General question on this issueMatt Craighead
                                                  General question on this issueKSP
                                                    General question on this issueMatt Craighead
                                                      Here's one...Singh, S.R.
                                                        Here's one...Anonymous
                                                          "peak output neutron cloud" Of what? (NT)Michael Menietti
                                                      General question on this issueKSP
                                          ia64 scheduling.. it is the point.john evans
                                    Loads and adds from a circuit perspectiveAnonymous
                                      You would prefer displacement then? (NT)David Kanter
                                      Loads and adds from a circuit perspectiveAnonymous
                      90nm A64 ReviewPeter Huisken
                        90nm A64 ReviewPaul DeMone
                          90nm A64 ReviewIntelUser2000
                      Why is I2 x86 logic box so big?IlleglWpns
                        Its small compared to Merced'sPaul DeMone
                90nm A64 ReviewMarc M.
              90nm A64 Reviewmas
      Prescott extra transistorsIntelUser
    90nm A64 ReviewPaul DeMone
      PossibleIlleglWpns
      90nm A64 Reviewsuperfunk
        90nm A64 ReviewPaul DeMone
          90nm A64 Reviewj
            Understandable mix upPaul DeMone
      90nm A64 ReviewPeter Gerassimoff
        90nm A64 ReviewIntelUser2000
          90nm A64 ReviewPeter Gerassimoff
          90nm A64 ReviewPeter Gerassimoff
  90nm A64 ReviewVincent Diepeveen
  Some 90nm A64 Power DataPeter Gerassimoff
    Some 90nm A64 Power DataAlberto
      Please clarify (NT)Dresdenboy
      Some 90nm A64 Power DataDresdenboy
      Some 90nm A64 Power DataPeter Gerassimoff
        Some 90nm A64 Power DataDavid Kanter
    Some 90nm A64 Power DataPaul DeMone
      Quality assuranceDoug SIebert
        Quality assurancePaul DeMone
    Some 90nm A64 Power DataIntelUser2000