Northwood vs. Prescott, an early hands-on look

Straight_ManStraight_Man Geeky, in my own wayNaples, FL Icrontian
edited February 2004 in Hardware
There is a Loyd Case article on Extremetech that folks thinking about Prescott might want to look at....

Link: http://www.extremetech.com/article2/0,3973,1478837,00.asp

Basicly, Intel is finding it easier to use the old architecture sizing to produce 3.4 GHz CPUs than to produce Prescotts in terms of yield volume. Prescott 3.4's will be alot rarer and more costly at official launch than Northwood and EE CPUs at same or close to same true speeds inside CPU. Also look at cache sizing on the new Northwoods and EEs when you decide what to get and how much to spend.

Due to rarity of 3.4 GHz Prescotts at launch and close to launch time, getting the latest highset speed Prescott real close to launch time might be something where you spend more money than what the cost justifies in terms of benefits gained, simply due to how many Prescotts that are stable at 3.4 come out of production-- for now and near future. Expect 3.4 GHz Prescotts to be VERY expensive, even after you take into account the increased benefits.

John D.

Comments

  • mmonninmmonnin Centreville, VA
    edited February 2004
    Prescott is just something to brdge the gap after Northwood. Seems to not be worth it to me.
  • rykoryko new york
    edited February 2004
    Yeah, the sweet spot right now is with the 2.8C's @ around $180 and falling on Newegg.com. Personally, I am going to wait until socket775 and all of the changes that go along with it like PCI-express, etc... to upgrade to a prescott.
  • CAGCAG Boca Raton, Florida
    edited February 2004
    mmonnin wrote:
    Prescott is just something to brdge the gap after Northwood. Seems to not be worth it to me.

    Not Prescott but the socket478 Prescott launched on 2/2. Wait 'till the LGA775 Prescotts launch. Then we'll see something. Could never figure out why Intel launched an S478 Prescott. Here's an article that seems to agree http://the-inquirer.com/?article=13996
  • Straight_ManStraight_Man Geeky, in my own way Naples, FL Icrontian
    edited February 2004
    CAG wrote:
    Not Prescott but the socket478 Prescott launched on 2/2. Wait 'till the LGA775 Prescotts launch. Then we'll see something. Could never figure out why Intel launched an S478 Prescott. Here's an article that seems to agree http://the-inquirer.com/?article=13996

    True, IF the yield rate problem can be solved very well. The very same thing that happened here can happen with the next gen, as it is even harder to get right reliably. They can make some, sure, and ahve in lab, but have not said exactly how many turn out good and how many have to be run slower, with is one of the points made in Loyd Case's article. Have you heard "rarer than a perfect diamond" used, ever??? It looks like one reason for the release for the Northwood type, is THOSE they could make more successfully overall, and ditto the EE subtype-- and could price them accordingly.

    However, there are two other major issues-- can they relaibly make a chipset and can other mfrs make things like video cards, that reliably can work with however many CPUs of the next gen are actually made successfully?? How long will that take?? Can RAM be made fast enough to take advantage of how well the new CPUs could use RAM??? A box with a fast CPU and slower rest of components is a box where if the CPU does not wait for the rest of the things needed to get info to and from user, the components fail from overload and malf. To get a whole gen of box needs lots of parts.

    AND we still do not know exactly why things failed as far as volume production at goal speed (Intel HAS told folks who have contacts at high levels that at launch time they will say there will be very limited amounts of Prescotts available at 3.4 GHz). For the first chips). AND we do not have Intel's word in writing yet that the Prescotts will or can be true 64 bit processors-- just hopeful rumours without official confirmation.

    All I am saying is wait in this case until we see evidence before pinning too much on having Prescott available in enough quantity to be economical for any but the richest, adn wait for Intel to fill in some ominous silences with facts, or wait until the actual performances are MEASURED in the field more than once by trusted folks who have proven track records to jump on the Prescott bandwagon.

    John D-- who only leaps on solely "a wing and a prayer" if he has no other choice or has enough knowledge of certian kinds that it is NOT "just a wing and a prayer." Icarus tried to fly to the sun on wings, but the material holding the wings together was WAX, which melted(the wings had a very fatal mfring defect in them)-- so the story goes. That story has a point, same one is made by kids who did drugs of very dangerous kinds, thought they could fly, tried, and died. I want to see that this very expensive looking thing will not die before half the warranty is up more than half the time before going to this particular gen myself. Color me skeptical, I have seen comparable things happen again and again and AGAIN in IT over the years.
Sign In or Register to comment.