1. This forum section is a read-only archive which contains old newsgroup posts. If you wish to post a query, please do so in one of our main forum sections (here). This way you will get a faster, better response from the members on Motherboard Point.

Intel is cutting off making the Prescott???

Discussion in 'Overclocking' started by dennis e strausser jr, Dec 31, 2004.

  1. < < < <Prescott is a
    < < < <runt, as evidenced by the fact that Intel have already canned the
    < < < <development on it.

    Really?

    What the hell is this ?

    http://processorfinder.intel.com/scripts/details.asp?sSpec=SL82U&ProcFam=483
    &PkgType=ALL&SysBusSpd=ALL&CorSpd=ALL

    We're already 200 MHz away from 4 GHz.
    Yet, ppl claim we are still hanging around the 3 GHz range.

    My guess is that there's already a 4 GHz chip @ prototype stage, and it's
    possible
    that there is something faster that's not far from the prototype stage..


    Temps, it's not all that new of a thing for a Pentium 4 or Pentium 4 Based
    CPU runs hot.
    When I get the parts I need for my Xeon's, the temps are listed on Intel's
    site as being about the
    same.
    Intel did do (Very, Very) well at keeping the Prescott's temps down.
    Compare 77 Watts = 72ºC for my Xeon's
    with 115 Watts for the Prescott and the same temps for both, or close to.
    More power, and a much smaller design, and Intel was somehow able to keep
    temps down that much..
    Another link. This one goes to the page with my Xeon's on it, and thermal
    temps.

    http://processorfinder.intel.com/scripts/details.asp?sSpec=SL73M&ProcFam=528
    &PkgType=ALL&SysBusSpd=ALL&CorSpd=ALL

    Now, before more ppl complain about the temps of the Prescott, keep this in
    mind.
    Intel somehow managed to keep the same temps even though the speed is a bit
    different,
    and it's power use.
    115 - 77 = 38 So that's already 38 Watts more then my Xeons.
    Next
    3800 - 2660 = 1140 MHz more then my Xeon's.
    So as for Thermal problems,,, What thermal problems.
    Last..



    This is a question, Thermal Spec
    72°C

    What does intel mean by this, the hottest, or under load?
    Hmm..
    I Just know they both have about the same Thermal Specs.
    Means Intel did a good job..

    Denny. ;-) :)

    Denny. ;-) :)
     
    dennis e strausser jr, Dec 31, 2004
    #1
    1. Advertisements

  2. From whats here, Intel is a little behind.
    The Gallatin - P4EE
    They had said they will still make some old 130nm line yet.
    But this is not on the market.
    Fsb = 1066 MHz or by a good guess 266 MHz FSB (266.5)
    The first Prescotts of 2005 are supposed to use this bus too.
    Before it gets replaced by the,,,, it's in here..

    http://www.a1-electronics.net/Intel_Section/CPUs/Roadmap.shtml

    Their not Scrapping the Prescott, it's just almost reached it's end.
    Q1 of 2005 will be the last new Prescott's, then SmithField will take it's
    place.

    This is, in a very real way, the same thing that happened when they first
    released the P4.
    It really wasn't that fast a chip, then they made some advances to it's
    core.
    100 MHz FSB or 400 MHz total bus was the same as a lot of the PIII's of that
    same time.
    And as a matter a fact, some of the P3's were faster then the P4's @ the
    time.


    Denny. ;-) :)
     
    dennis e strausser jr, Dec 31, 2004
    #2
    1. Advertisements

  3. dennis e strausser jr

    Noozer Guest

    http://processorfinder.intel.com/scripts/details.asp?sSpec=SL82U&ProcFam=483
    How many people want a CPU that could be also used to make coffee and risks
    burning your house down?

    Prescott is a piece of crap. All Intel is doing is trying to push more
    electricity through it to make it go faster, not design a better chip.
     
    Noozer, Dec 31, 2004
    #3
  4. dennis e strausser jr

    Noozer Guest

    Prescott is hardly a year old and it's already over? Sounds like sCRAPping
    to me. Prescott should never have seen the light of day.
     
    Noozer, Dec 31, 2004
    #4
  5. There's less risk of fire with a Prescott then there is a space heater.

    Here's the links one more time.
    Prescott Thermal Specs first.
    http://processorfinder.intel.com/scripts/details.asp?sSpec=SL82U&ProcFam=483

    http://processorfinder.intel.com/scripts/details.asp?sSpec=SL73M&ProcFam=528

    Hm.
    72ºC for the Intel Xeon 2.66 Thermal Spec.
    72.8ºC for the 3.8 GHz Prescott Thermal Spec.

    You tell me, where did I miss something here.
    From what it looks like to me is, Intel was able to pump more power on one
    side
    of the CPU, up the speed a bit, but still maintain the same thermal specs
    of that of a lesser
    power CPU.

    Hm??
    So if I overclock my Xeon's high enough to add vcore, I might just run
    hotter then a Prescott.
    Seems very true..
    Denny. ;-) :)
     
    dennis e strausser jr, Dec 31, 2004
    #5
  6. Yeah & yet another reply.
    Oops, I goofed on something.that
    same time. < <
    They still used the x2
    They were 100 & 133 x 2. PII & PIII..
     
    dennis e strausser jr, Dec 31, 2004
    #6
  7. Yes Dennis, didn't it sink in first time round? Prescott is a screw-up and
    Intel have canned development on it, as well as its 65 nanometre sucessor
    Tejas.

    By the way, is English not your first language, or are you deliberately
    trying to twist/misrepresent what I said in the previous thread? At no point
    did I claim that Intel were "cutting off making" them. They haven't got any
    choice but to continue making Prescotts right now, as they don't have
    anything else to sell in the desktop space.
    Erm, a 3.8GHz 1MB Prescott. And?
    Your point being? If the inference is that Intel will have 4GHz (and faster)
    P4's out soon, you're entirely incorrect.
    It wasn't actually me who said that, but the point you seem to have entirely
    missed is that CPU's have been sat in the "three point something" GHz range
    for over a year - when Intel's roadmap from the beginning of 2003 said that
    Prescott should be at 5.2GHz by now.

    In other words, even by Intel's own standards, they've failed to achieve the
    ramps in clock speed they previously expected to, basically because, as
    mentioned, Prescott is flawed and has failed to scale in the way it was
    designed to.
    Do you not follow technology news at all? Intel did have a prototype 4GHz
    CPU, but they couldn't get it to work at productionable yields, so they
    cancelled it back in the summer. The rest of the world knew this months ago
    so go check your facts.
    Lol, no, there isn't. Since their failure to get the 4GHz CPU to work, Intel
    said quite publicly that Prescott will not go any faster than 3.8GHz in
    retail form. This is part of the reason why they invented their new CPU
    numbering scheme - they couldn't market on GHz alone anymore.

    The next line of CPU's, the 600 series, will be available from 3.2 to
    3.6GHz, but with 2MB of L2 cache and EM64T extensions to try and pep up the
    performance, while after that the "Smithfield" dual core CPU will be
    available at speeds between 2.0 and 2.8GHz.

    Note though that Smithfield (basically two Prescotts in a single package
    with 1MB shared L2 cache) is no more than a stopgap measure to try and keep
    a nominal presence in the top-end of the market until they can bring the
    "Merom" Pentium-M based processor to production - at the end of '05 or early
    2006.
    What planet are you living on? Prescott runs a lot hotter than Northwood at
    the same core speed, and that's pretty widely accepted on Earth.
    Fook, if "your Xeons" run that hot, you should probably check the heatsink
    installation. I have to ask though why you are using a server level CPU
    based on 0.13 micron architecture as a comparison for a desktop chip.
    That's part of the problem. It's Prescott's inefficiency that causes the
    high thermal design power, which leads to the huge CPU heat output, and in
    turn places a much greater load on the motherboard's power supply
    components, and consequently on the PSU itself.
    Rofl, you're having a laugh aren't you? 115W TDP isn't a thermal problem?
    You haven't got a clue what you're talking about Dennis. This much is
    obvious from the ignorance you demonstrated further up the thread. You're
    one of the very few people who think Intel did a good job with Prescott.
    They f*cked it up, and they know it better than anyone.


    Richard Hopkins
    Cardiff, Wales, United Kingdom
    (replace .nospam with .com in reply address)

    The UK's leading technology reseller www.dabs.com
    Get the most out of your digital photos www.dabsxpose.com
     
    Richard Hopkins, Dec 31, 2004
    #7
  8. I give, I'm fighting a lost cause.
    Why, I don't know.
    I just guess I'm still hanging on to Intel, even though I like (You know who
    "AMD" Better. )
    It's almost like I think something will happen, even though I know it won't.
    And I also have a feeling that it won't be AMD or Intel that will take the
    market...
    It will be Apple..
    Don't know why I thnik this, it's just a feeling.
    In other words,,,,, ur right.... :)
    Yeah well....

    Denny. ;-) :)
     
    dennis e strausser jr, Jan 1, 2005
    #8
  9. That's what you get for talking bullshit on Usenet. Someone (or in this case
    a few people) will always call you on it.
    I was wondering this myself. I did wonder whether you were trolling but the
    attempted justification is far too straight-faced and serious to be anything
    other than sincere.
    Eh? Could you reformat the paragraph above so it makes sense in English?
    Apple taking the market? Are you on drugs?

    BTW...
    Lol, you don't learn do you? No matter how you try and mince words it hasn't
    done its job and *is* being scrapped/retired early. They're not developing
    any new steppings for Prescott and are aborting Tejas, its successor,
    altogether. Intel originally thought this CPU design would last them over
    three years. It will in reality last about a third as long.

    Rather than launching Prescotts with higher and higher clock speeds before
    transitioning to Tejas and the 65nm process as per their original intention,
    they've been stuck at more or less the same speeds as they were with
    Northwood, and been forced to bolt extras like 64 bit extensions and even
    more L2 cache on to try and keep their "new" product line fresh, inventing
    the product numbering scheme in a very rushed and haphazard way to disguise
    the fact that they can't sell on gigahertz anymore.

    This turn of events is a major downer for Intel as not only do they lose
    more of the gaming/top-end desktop market to AMD, more cache means even
    higher power consumption, while the higher component count means larger dies
    and lower yields, both of which greatly increase the unit production cost.

    Don't forget that the original Prescotts only got 1MB of L2 in the first
    place because Intel had to do something about the performance disparity with
    Northwood (and by implication AMD).
    Smithfield (small f) is just two Prescotts in a single package with 1MB of
    shared L2 cache. It is no more than a hastily designed, emergency
    replacement, not the "real" performance CPU that Intel had originally
    planned. Prescott's replacement was supposed to be Tejas, which is now dead.
    No, it's feck all like what happened when the Willamette P4 first came out
    (and was then replaced by Northwood). It's true that Willamette was a
    failure and that Northwood was a vast improvement. If Tejas had performed
    according to expectation then the same situation may have repeated, but
    Smithfield is not Tejas.

    Putting two crap, power-hungry cores on a single package is not the solution
    Intel wanted, but given the timescales, it doesn't have much option. It
    needs to bring out something "new", as failure to do so would slow the
    entire PC market and lose even more share to AMD. It will only be around for
    as long as it takes to engineer and produce Merom.
    Lol, you're really stupid! Even the original P4's system bus is *waaaay*
    faster than anything the Pentium!!! had. The 133MHz Pentium!!! bus ran at
    1.06GB/sec (and was hitting the buffers), while the P4's started at
    3.2GB/sec and had scalability designed in.
    Yeah, they were, but that was mainly because the P4's had much longer
    pipelines than the Pentium !!!'s. The lengthened pipeline was implemented to
    enable the P4 core to scale to much higher clock speeds, but the downside
    was that, clock for clock, a Pentium!!! would have better performance.
    Exactly the same situation pertains when comparing Northwood (20 stage
    pipeline) with Prescott (31 stages). If Prescott had gone to 6GHz and Tejas
    to 10, as per their original design briefs, nobody would have cared.
    So why did you bother then? It's difficult to put this any other way Dennis,
    but you come across as an ignorant idiot without any real grasp of the
    subject you're trying to discuss. If you're not trying to troll, maybe you'd
    be better off keeping away from/not starting threads like this huh?
    --


    Richard Hopkins
    Cardiff, Wales, United Kingdom
    (replace .nospam with .com in reply address)

    The UK's leading technology reseller www.dabs.com
    Get the most out of your digital photos www.dabsxpose.com
     
    Richard Hopkins, Jan 1, 2005
    #9
  10. In other words for now, I might as well just stick with AMD, or older CPU's
    by Intel.

    Question that's a lot more of this group.
    2.66 GHz how far over do you think I may be able to go?
    I'm guessing around 3.2 or so. I have air cooling for now,
    DDR266 mixed with DDR333, only for 1 GB.
    I'm planing on getting 1 Gig Dimms, but it will be a while.
    Board is about the only one I know of that will allow overclocking.
    If you know of any besides Asus "PC-DL Deluxe" Please tell me.
    Ty for all advice and help in correcting me.

    Denny. ;-) :)
     
    dennis e strausser jr, Jan 1, 2005
    #10
  11. dennis e strausser jr

    Spajky Guest

    real life performance has not much to do with theorethical max
    bandwith on the same real clock, since a waste majority of
    calculations pass thru chipset & memory & real life performance of it
    matters (not its theoretical max bandwith); how certain chipsets &
    memory really perform per clock; see my site under comp/benches (first
    "new") ... sure, higher real clock better performance .. :)
    yes, the efficiency dropped with that marketing pumping up clock; now
    happens contrary (I mean P-M) - enhancing the previous P-3 core with
    some P4 features (peripherals) .. you can see in % per clock
    efficiency on the same my site, but second "new" .. :)

    Happy New Year, Richard!
     
    Spajky, Jan 3, 2005
    #11
  12. dennis e strausser jr

    Minotaur Guest

    Apple, never LOL they don't even have an e-mail client capable of
    connecting to an Exchange Server correctly.

    Intels problem in the future shall be the choke on the system bus, from
    multicore processors.

    AMD is it! Intel might make a come back when the Marketing dept isn't in
    charge of CPU development. Server sales of Opterons are going through
    the roof, while Itanium sales sit idle. Not worth mentiong Apple, they
    have no 'real' server sollutions and never shall.
     
    Minotaur, Jan 3, 2005
    #12
  13. dennis e strausser jr

    Immuno Guest

    Hang on a mo' - can we get the terms of reference of this "Pissing
    Contest" - namely alt.comp.hardware.overclocking back into focus.

    Correct me, but underlying it all is getting the biggest bag for your
    buck/quid/eurineo...

    So: (personally) I don't give a flying fart how many of who's GHz it does -
    its results that matter - isn't it?

    Much against my better judgement I got blagged into delving into 939
    territory by my local clan of the AYBABTU/CS2 mob. So we have (burning in
    behind me as I write):

    939 Winchester 3000+
    Asus A8V ver2.0
    2 x generic DDR400
    300 Watt PSU (Ebuyer) !!! :eek:)

    On stock air with a little tweaking and squeezing the ram timings a little:

    CPU-Z - HTT 280MHz and 2520MHz
    Super-Pi - 34s
    Everest - Memory read @ 6.060Ghz and latency @ 44.7ns
    32 degC at rest and 50 degC folding

    ..... OK so the rig still gets smoked by a whole stack of P4 whatevers - but
    not for only ~200 hundred quid all in it don't!!

    :eek:)

    Pete

    For further data I need to negotiate about getting my heavily modded 9700pro
    back out of my son's box!
     
    Immuno, Jan 4, 2005
    #13
  14. dennis e strausser jr

    No_ONE_Here Guest

     
    No_ONE_Here, Jan 4, 2005
    #14
  15. dennis e strausser jr

    Immuno Guest

    Yeah, I know - your quite right. It just started to get a bit
    "introspective". :eek:)

    Nice O/C on the Xeon tho'

    Pete
     
    Immuno, Jan 4, 2005
    #15
  16. dennis e strausser jr

    Frank Guest

    POS? Not even close. I'm running a Prescott 2.8 oc'ed to 3.7 air cooled
    idle temp 41c on a P4C800ED.
    Frank
     
    Frank, Jan 31, 2005
    #16
  17. dennis e strausser jr

    Thomas Guest

    Idle doesn't say much, what's your stressed temp?
     
    Thomas, Jan 31, 2005
    #17
  18. dennis e strausser jr

    Frank Guest

    Highest I've seen is 122F. And yes idle does say a lot.
    Frank
     
    Frank, Jan 31, 2005
    #18
  19. dennis e strausser jr

    Noozer Guest

    All idle says is that you spent $100+ on case fans and your PC sounds like a
    jet.

    My 2.6Ghz Northwood running at 3.2Ghz idles at 24'C and is dead silent
    (except for the HDD). Under load it might hit 40'C. It can run Prime95 for
    DAYS without errors.

    How long will your machine run Prime 95 without reporting any errors - and
    at what temp?
     
    Noozer, Jan 31, 2005
    #19
  20. dennis e strausser jr

    Frank Guest

    "$100 on a case fans"...never, not in mt life time! "Sounds like a
    jet"...not any of my six!
    What planet are you from and how long have you been oc'in?
    You sound like a newibe or a blind squirrel who just happened to get
    lucky and find an acorn.
    Frank
     
    Frank, Jan 31, 2005
    #20
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.