1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Intel Pentium Extreme Edition 955

Discussion in 'Article Discussion' started by Tim S, 27 Dec 2005.

  1. -EVRE-

    -EVRE- New Member

    Joined:
    12 May 2004
    Posts:
    372
    Likes Received:
    1
    what about the socket 754? they ditched that after the 3700+ with no plans to further the socket. That makes me want to cry. Im still hope they release a dual core sempron for it.

    BTW. I do have a socket 939 Opteron 170 (for those that dont know what that is, its: dualcore 2.0ghz 1mb cash per core) and WOW does it fly! havnt overclocked it yet, its still in burn in.
     
  2. Darkedge

    Darkedge Member

    Joined:
    26 Nov 2004
    Posts:
    363
    Likes Received:
    0
    "Thanks to Intel's Flex Memory technology, it's also possible to install differing memory sizes in each slot while still utilising in dual channel mode. This is great, as it allows a flexible upgrade path, as you could have 1GB of DDR2 installed now, and you could add two 1GB modules for a total of 3GB of memory in dual channel."

    err am I the only person who is a little confused about this bit? Dual channel cannot run surely though THREE slots on the mainboard - more accurately would be saying you could have a 1GB stick and add a 512 and use Dual channel access..
     
  3. Tim S

    Tim S OG

    Joined:
    8 Nov 2001
    Posts:
    18,881
    Likes Received:
    78
    No, you've misunderstood what I wrote. 1GB of memory (as in 2x512MB) installed now, and add another 1GB modules (i.e. 2GB) for 3GB total filling all four slots.
     
  4. Darkedge

    Darkedge Member

    Joined:
    26 Nov 2004
    Posts:
    363
    Likes Received:
    0
    ahhh I see - that makes much more sense.
     
  5. Tim S

    Tim S OG

    Joined:
    8 Nov 2001
    Posts:
    18,881
    Likes Received:
    78
    :thumb:
     
  6. hitman012

    hitman012 Active Member

    Joined:
    6 May 2005
    Posts:
    4,877
    Likes Received:
    19
    Although I'm not their biggest fan, I have to give credit to Intel for the double-core idea. They don't have to make sure both cores pass at, say, 3.2GHz and bin if they don't - they can simply take the core that doesn't make it, bin that, and pair it with another 3.2GHz core. On top of that, the two dies are both much smaller and therefore (I assume) easier to produce - sharing the Cedar Mill design with the single-cores is nifty.

    I think that Intel are coming back on form, although I don't feel they can really best AMD with much until they phase out NetBurst and focus on their upcoming Dothan-based core. I'm very much looking forward to Conroe... only a year or so to go :naughty:. That having been said, the 65nm process is helping them actually hit a higher clock speed than 3.2GHz, which must be a relief for them as they're currently being thrashed on power efficiency. What's the TDP of the 955?
     
  7. Tim S

    Tim S OG

    Joined:
    8 Nov 2001
    Posts:
    18,881
    Likes Received:
    78
    130W IIRC
     
  8. atomo

    atomo New Member

    Joined:
    28 Dec 2005
    Posts:
    5
    Likes Received:
    0
    intel sees the future of gaming

    I feel this review is a little nearsighted. It definitely covers the "now" factor, but is not looking to the future of gaming. Intel says they are targeting the hardcore gamer, and it looks to be exactly what they are doing.

    Although AMD clearly has a foothold on performance, something I think a lot of people are overlooking is the bigger picture of gaming: multicore cpus and multithreaded games. Consider that all the next-genereation consoles have multi-core processors. This forces developers to build multi-threaded games. I believe the thinking form intel is that when multi-threaded games are the norm, and developers are comfortable working in that environment, we're going to see a whole lot more performance from the intel camp. They will be ready and waiting to throw more threads down than a wild sewing machine.

    There is still a good chance that AMD will stomp all over them in that arena too, particularly if intel doens't get over their P4 architecture (completely), but only real-world tests will tell.

    I for one, have been waiting for true multi-cpu, multi-threaded, multi-whatever games ever since I got my hands on a dual cpu system about 5 years ago. :D
     
  9. specofdust

    specofdust Banned

    Joined:
    26 Feb 2005
    Posts:
    9,568
    Likes Received:
    168
    Good review Tim, I am curious though, why are Intel bothering, with Yonah so near round the corner? From what I know of chip architecture, its damn expensive to get a new chip going, I know this is just, basicly, an updated dual core prescot, but still, why bother when we've got dual core dothan type CPU's going at about 2.2-2.4(from what I hear) coming within months, that will obviously be better processors for most of the CPU market?
     
  10. FIBRE+

    FIBRE+ Active Member

    Joined:
    27 Feb 2005
    Posts:
    1,597
    Likes Received:
    11
    Ah ok, was thinking of the motherboard reviews as a reference as they've had 7800's in, I also presumed it was a review rather than a preview :duh:

    :)
     
  11. Tim S

    Tim S OG

    Joined:
    8 Nov 2001
    Posts:
    18,881
    Likes Received:
    78
    Hi and welcome to the forums.

    Thanks for stopping by to comment.

    I do agree with what you're saying, which is why we've not finished looking at the performance of this (and other) Presler CPU(s). AMD has an announcement next month, and it would be suitable to start looking at things a little deeper at that time, or maybe after that review is published.

    There's a lot to cover (as is stated during our results analysis) with respect to 'the future' and there's no doubt we'll be looking at it. The future is important to a lot of people here, because they can't keep dropping thousands of pounds on computer equipment everytime Intel/AMD/ATI/NVIDIA release something new. We'll attempt to answer these questions regarding the future as best we can when we continue investigating Intel's latest technology in the new year. :)
     
  12. hitman012

    hitman012 Active Member

    Joined:
    6 May 2005
    Posts:
    4,877
    Likes Received:
    19
    Not bad at all considering that it's roughly the same as the 3.2 Smithfield. I suppose the real gains come when the newer cores are made on the 65nm process. It'll be fab (sorry, that was such a terrible pun).

    I'll be very interested to see what AMD come up with against Conroe et al. One thing is for sure - it's going to be a lot tighter than P4 vs. A64.
     
  13. Nature

    Nature Member

    Joined:
    21 Nov 2005
    Posts:
    492
    Likes Received:
    1

    My question in relation to yours pig>: Why would Intel direct they're marketing of gamer processor's here? The Pentium D 780 almost kicks everythings ass in games (especially when overclocked to competing processor speeds. The Yonah is a double Pentium D with a higher FSB, clock speed, quite possibly dual channel memory support, and in late 06' Intel is touting a 4mb on die cache (2mb per core).

    What the fork Intel? Do you look at your own Benchmarks>!?
     
  14. Stephen Brooks

    Stephen Brooks New Member

    Joined:
    24 Apr 2005
    Posts:
    40
    Likes Received:
    0
    ^^^ Replace "D" by "M" in the above post?

    "Much"? I'm almost of the opinion 3.46 is a decidedly unimpressive speed for this core given how far it reportedly overclocks. :)
    I imagine if there hadn't been a Conroe or whatever coming down the pipeline, they could have done a 3.73 version of this chip easily, but don't want to step on the toes of the new chip.

    Or, to put it another way, there were 3.4GHz chips released on the 130nm process. That's now TWO generations ago, albeit not dual core, but you see where I'm coming from. If they expected even a 20% speed improvement per generation (which was the sort of thing everyone was bandying about back in 2001), we'd now be at 4.8GHz or so... (c.f. Cedar Mill overclocking?)

    On the other other hand, 2*3.46 = 6.93 > 4.8 so if you're easily parallel you're winning.
     
    Last edited: 29 Dec 2005
  15. hitman012

    hitman012 Active Member

    Joined:
    6 May 2005
    Posts:
    4,877
    Likes Received:
    19
    Very true - the problem is keeping heat output levels within an acceptable range. Smithfield delivered a record-breaking (for the desktop) 130W TDP, and this hasn't improved on that - it's just achieved a higher speed with roughly the same output.

    It might overclock very nicely on a high-end watercooling system, but obviously this isn't feasible easily, even if this chip is aimed at the gaming crowd. Overclocking it to 4.2GHz, assuming a voltage of 1.55V, would require over 200 watts at peak load. You can run a whole system off that!

    The real benefits of this 65nm process are coming with Merom and Conroe in Q4 '06, and for that... can't wait :D
     
  16. Tim S

    Tim S OG

    Joined:
    8 Nov 2001
    Posts:
    18,881
    Likes Received:
    78
    It's heat dissipation remains the same (TDP), but the power consumption has dropped considerably. Hitman, I think you're confusing TDP with power draw. TDP is the Thermal Design Power, not the power consumption.

    Check here for some power consumption numbers - the text is a little confusing IMO, as they've mixed power consumption with heat dissipation (wasted heat/thermal output): http://www.xbitlabs.com/articles/cpu/display/presler_8.html

    The XE 955 consumes as much power as a single cored P4 XE 3.73GHz and less than the XE 840 and P4 670. :)
     
  17. Hamish

    Hamish New Member

    Joined:
    25 Nov 2002
    Posts:
    3,649
    Likes Received:
    4
  18. hitman012

    hitman012 Active Member

    Joined:
    6 May 2005
    Posts:
    4,877
    Likes Received:
    19
    Sorry if it wasn't clear in my post... I was referring to the thermal output in the first bit and the consumption in the second - rather contradictory, really.

    The large TDP was, as far as I know, the most important factor limiting them from creating faster models of Smithfield; dissipating 125W on something the size of your thumbnail is asking a lot of a heatsink. The shrink has improved matters, with more cache, higher speeds and lower power consumption, yet the high levels of heat dissipation when compared to any AMD offerings still remain.

    That having been said, they're going in the right direction now - kooking at those charts, it is impressive how they have managed to drop it. 170W for the 840 XE! And this is faster, too :)
     
    Last edited: 29 Dec 2005
  19. Tim S

    Tim S OG

    Joined:
    8 Nov 2001
    Posts:
    18,881
    Likes Received:
    78
    Yes, it's quite crazy... but, it's good to see that they're heading in the right direction.
     
  20. Highland3r

    Highland3r New Member

    Joined:
    25 Jul 2003
    Posts:
    7,553
    Likes Received:
    16
    Yonah is a mobile/laptop based chip, its not intended as such for desktop machines..
    Its still gonna suck at certain things as Dothan does atm but the 2 cores will make it a much much more useable CPU imo.

    The motherboards lined up don't seem to shabby either, we should see some decent desktop boards out (afaik Asus are developing a Crossfire based board...) for the chips, making them a viable option for Desktop based systems...
     
Tags: Add Tags

Share This Page