1. OCAU Merchandise is available! Check out our 20th Anniversary Mugs, Classic Logo Shirts and much more! Discussion in this thread.
    Dismiss Notice

Intel “Paxville” Dual Core Xeon tested

Discussion in 'Intel x86 CPUs and chipsets' started by Tony, Oct 20, 2005.

  1. Tony

    Tony New Member

    Joined:
    Jun 26, 2001
    Messages:
    9,987
    Location:
    Sydney, NSW, Australia
  2. dinos22

    dinos22 Member

    Joined:
    May 8, 2004
    Messages:
    12,332
    it gets breaten convincingly but it's realy days for that platform........i'm sure there is more to it
     
  3. bsbozzy

    bsbozzy Member

    Joined:
    Nov 11, 2003
    Messages:
    3,924
    Location:
    Sydney
    All i can say is Ouch Intel.
     
  4. lazyboy1984

    lazyboy1984 Member

    Joined:
    May 30, 2004
    Messages:
    2,236
    Location:
    SE Melbourne
    now, who didn't see that coming anyway? putting a xeon tag on a smithfield can't make it go faster.
     
  5. -=N0N@ME420=-

    -=N0N@ME420=- Member

    Joined:
    May 9, 2004
    Messages:
    7,562
    I'd like to see chainbolts reaction :leet:
     
  6. V-Spec

    V-Spec Member

    Joined:
    Jul 14, 2003
    Messages:
    815
    Location:
    Melbourne
    exatcly!
    now when intel steers away from netburst. BAM they will own
    the should do it allready
     
    Last edited: Oct 20, 2005
  7. phextwin

    phextwin Member

    Joined:
    Aug 30, 2003
    Messages:
    4,221
    Location:
    Geelong Velocity:Unknown
    Anyone know what the slot is on that board at the very bottom?
     
  8. bsbozzy

    bsbozzy Member

    Joined:
    Nov 11, 2003
    Messages:
    3,924
    Location:
    Sydney
    No doubt he will winge at me ;)
     
  9. -=N0N@ME420=-

    -=N0N@ME420=- Member

    Joined:
    May 9, 2004
    Messages:
    7,562
    http://www.cdrinfo.com/Sections/News/Details.aspx?NewsId=15243

    and from the article:

    http://www.gamepc.com/labs/view_content.asp?id=paxville&page=4



    Actually I've seen him in this thread several times, no doubt hes staying well clear of it. :p
     
    Last edited: Oct 20, 2005
  10. chainbolt

    chainbolt Member

    Joined:
    Jun 27, 2001
    Messages:
    113,221
    Location:
    Tokyo_Munich
     
  11. FaTs

    FaTs Member

    Joined:
    Aug 25, 2002
    Messages:
    1,455
    Location:
    Bris Vegas...
    Lots of people run DP Xeons I don't think there is much of an end user market for MPs....
     
  12. chainbolt

    chainbolt Member

    Joined:
    Jun 27, 2001
    Messages:
    113,221
    Location:
    Tokyo_Munich

    Exactly, therefore my hint at Dempsey. I still have to meet an OCAU member that runs Xeon MPs. And I wonder therefore, just a little, about the sudden great attention of desktop AMD users into an Intel MP product like Paxville. :lol: :lol: :lol:

    But I am sure there is good reason. ;) :leet:
     
    Last edited: Oct 21, 2005
  13. chainbolt

    chainbolt Member

    Joined:
    Jun 27, 2001
    Messages:
    113,221
    Location:
    Tokyo_Munich
    No, I don't care to make predictions about Dempsey's performance. I have never run a DP Xeon, and I have no intention to do so - but I am sure our AMD friends in this thread know a lot about this.

    The reason why Dempesy is supposed to perform better and run cooler is that it is based on 65nm technology and the Presler core. Furthermore it will use the upcoming Bensley platform which seems to be superior to the Lindhurst ( E7520) platform they used for the linked Paxville test. Dempsey will attach each processor with a separate Front Side Bus (1066 Mhz), in case of Lindhurst the bandwidth is shared between the two processors and anyway only based on (800 Mhz).

    But I really don't care. I do not run Xeons be it MP or DP, and both products (pax and demps) will be replaced already next year at this time by entirely different products.
     
    Last edited: Oct 21, 2005
  14. cvidler

    cvidler Member

    Joined:
    Jun 29, 2001
    Messages:
    18,845
    Location:
    Canberra
    4-way P3 Xeons count? (they never got the 'MP' tag, but you needed them for 4-way, as regular P3's only went 2-way)

    But, yeah both AMD and Intel price there 'Xeon MP' and 'Opteron 8xx' range outside of what any home user is bound to spend.

    But I've personally had two quad boxes over the years, a quad PPro200, and a Quad P3 Xeon 700. But I'm sticking to my DP Xeons now (2x 3.06 and 2x 2.4). /me points to my site - www.datamine.tk for Xeon news (specifically ASUS related)
     
  15. chainbolt

    chainbolt Member

    Joined:
    Jun 27, 2001
    Messages:
    113,221
    Location:
    Tokyo_Munich

    Interesting. How did they perform in comparison to a regualr single processor deskotp system?

    If you are sticking to DP Xeons, you might be really interested in Dempsey and this new platform "Bensley". They should perform well together with 2 separated front sides bus.
     
  16. cvidler

    cvidler Member

    Joined:
    Jun 29, 2001
    Messages:
    18,845
    Location:
    Canberra
    I must admit I'm surprised you haven't gone SMP before. Try it, you'll like it.

    My main reasons for SMP are thus:

    - multi tasking, I have ALOT of software running all the time on my PC and servers, so SMP (be it, multi processor, multi core, and even hyperthreading) helps in day to day tasks, that few benchmarks can actually show it's something you have to try first hand.
    - Longevity, I find my single CPU mahcines need upgrading every 6-12 months, I keep my SMP boxes for upto 3 years before they're outside of my useability range.

    I've gotten into digital TV big a few years ago, and have dual tuners in my main PC (the 2x3.06 @ 3.8), the benefit of that is, I can (trans/en)code the resulting MPEG2 streams into xvid on the fly with no detrimental performance loss even while gaming (multiple HDDs and ample RAM are essential for that too).

    I've only just 'moved on' my two dual P3 boxes, more a reason of space, than not being usable, but they were over 3 years old and still capable of doing worthwhile work as a PC or server. One ended up a family a hand-me-down (whihc still plays recent games well with it's 9700), the other was sold off as a family PC.

    I forsee my two current Xeon boxes lasting me until Intel does move away from NetBurst, when once again I'll be buying in pairs (dual core or not). Or if AMD's Opteron prices come down enough, I maybe buying a pair or two of them.

    edit: as for a comparo, I haven't had a UP machine in many years. Cyrix M2 233 mean anything to anyone?
     
    Last edited: Oct 21, 2005
  17. chainbolt

    chainbolt Member

    Joined:
    Jun 27, 2001
    Messages:
    113,221
    Location:
    Tokyo_Munich
    I am sure I would. But I am changing my hardware all 2 or 3 months. The one time investment into SMP for CPUs and the platform is simply too much for me.
     
  18. cvidler

    cvidler Member

    Joined:
    Jun 29, 2001
    Messages:
    18,845
    Location:
    Canberra
    You'll just have to sell a few more Porches next month then.

    In all seriousness, they're not that much more expensive. And if they last two or three times as long, then you've made your money back in productivity and longevity.
     
  19. -=N0N@ME420=-

    -=N0N@ME420=- Member

    Joined:
    May 9, 2004
    Messages:
    7,562
    I still am having trouble understanding the whole 65nm thing :(

    I thought the whole point of 65nm was mainly for financial gain, like you said, look at AMD, they're current 90nm tech is performing very cool and very fast, what is it about 65nm that is going to bring in cooler and faster? or is mainly 65nm + Presler core = more proft/performance increase + cooler?
     
    Last edited: Oct 21, 2005
  20. cvidler

    cvidler Member

    Joined:
    Jun 29, 2001
    Messages:
    18,845
    Location:
    Canberra
    90nm has been an embarrasment for Intel. Either they've royally messed up their new 90nm process (ie. manufacturing is bad), or they've pushed NetBurst far beyond what it is capable of as a design.

    65nm for Intel at least, will usher in a new manufacutring process, AND a new core design. They're hoping to fix all their current issues with this. Although it'll make it hard for anyone outside of Intel to truely know, if 90nm was a physical failure or NetBurst was beyond it's capabilities.

    but 65nm will mean.

    More chips per wafer = more profit (that's an easy one)
    New design = more performance (hopefully)
    New manufacturing process = power savings (hopefully)
     

Share This Page

Advertisement: