Intel's Dual-Core CPUs To Set New Record High In Power Consumption

KingFishKingFish
edited January 2005 in Science & Tech
Intel's upcoming dual-core processors once again will test the limits of power consumption. According to documents seen by Tom's Hardware Guide, the Smithfield CPUs are rated at a thermal design power of 130 watts, an increase of 13 percent from today's Prescott processors.
Dual-core and multicore chips promise to be one of the most important advances in processor development history. Intel and AMD claim to be able to achieve new performance levels by integrating two processor cores into one package. This apparently will be possible even with processor frequencies significantly below today's fastest processors. We were also told in the past that these speed gains will require less or little more power than an Athlon 64 or a Pentium 4 5xx/6xx.

At least Intel appears to miss this goal. Documents released to system builders specify the Thermal design power (TDP) of Smithfield processors at 130 watts. This represents an increase of more than 13 percent over today's Pentium 4 5xx (Prescott) and the upcoming 6xx (2 MByte L2 Cache), which post 115 watts. Maximum supply current climbs from 119 ampere to 125 ampere. The new chips also consume more power than Pentium 4 Extreme Edition 3.46 GHz processor (116.7 watts) and Intel's most demanding chip: The Itanium 2 1.6 GHz consumes 122 watts.
Source: Tom's Hardware Guide

Comments

  • GargGarg Purveyor of Lincoln Nightmares Icrontian
    edited January 2005
    A computer with a Smithfield CPU is just begging for an Eazy-Bake Oven® case mod.
  • KometeKomete Member
    edited January 2005
    If toms says 13% now it'll probably be 25% when they are released... and I'd like to see that Eazy-Bake Oven® case mod..hahha
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited January 2005
    Intel's TDP ratings are worthless. It's gonna draw 125a. If it's running at 1.5v, the REAL maximum heat output will be the better part of 187.5w...
  • LeonardoLeonardo Wake up and smell the glaciers Eagle River, Alaska Icrontian
    edited January 2005
    I'm waiting for Intel to spin this as something positive - that heat is a "sign of true, dual power!". Let's watch AMD. If there's no comment from their camp, it probalby means that AMD is expecting a heat problem as well with dual core offerings.
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited January 2005
    2 identical cores is gonna get you roughly twice the heat output of one of those cores. That's simply logical. Based on that, a dual-core opteron will probably run about 150w
  • LeonardoLeonardo Wake up and smell the glaciers Eagle River, Alaska Icrontian
    edited January 2005
    OK, so if you have both an AMD and an Intel dual rig system, you can have bacon and eggs for breakfast without a range top or hot plate. This is a great idea for dorm rooms.
  • Geeky1Geeky1
    looks at dual amd and intel systems next to him in dorm room...

    You don't need dual cores for that. They get hot enough as it is. :p University of the Pacific (Stockton, CA, USA)
    edited January 2005
    looks at dual amd and intel systems next to him in dorm room...

    You don't need dual cores for that. They get hot enough as it is. :p;D
  • edcentricedcentric near Milwaukee, Wisconsin Icrontian
    edited January 2005
    Intel will set the record for heat output.
    AMD will set the record for math output.
  • muddocktormuddocktor
    edited January 2005
    Geeky1 wrote:
    2 identical cores is gonna get you roughly twice the heat output of one of those cores. That's simply logical. Based on that, a dual-core opteron will probably run about 150w

    I don't think that AMD's offerings in dual core will be quite that bad. They (and IBM) are working on a new strained silicon process to be included along with SOI in future procs which is supposed to lower thermal output significantly over present 90nm A64/Opteron.

    Also, I seem to remember that Intel's published TDP numbers are derated from the actual max TDP of the P4 by around 15%, since they don't think that anyone or any program can keep the absurdly long pipelines in the P4 architecture full and processing 100% on the time. :rolleyes: I guess they've never heard of running 2 instances of folding on a HT-enabled P4. :rolleyes:
  • mmonninmmonnin Centreville, VA
    edited January 2005
    Ive heard it takes more than 2 Instances to keep an HT P4 full. Someone tested it and got slight improvement with like 4-5 WUs all at once. That really slows down the overall production of each WU but the PPW went up slightly with more than 2 WUs.
  • ThraxThrax 🐌 Austin, TX Icrontian
    edited January 2005
    Speaking of strained silicon, it's rumoured that the fabled SSE3-enabled E0 stepping from AMD is going to be arriving by Q205 with sSOI fabrication.
  • SputnikSputnik Worcester, MA
    edited January 2005
    and you thought your laptop ran hot now.... slap one of these babies in there and watch your pants light up!
  • RWBRWB Icrontian
    edited January 2005
    Sputnik wrote:
    and you thought your laptop ran hot now.... slap one of these babies in there and watch your pants light up!

    Been there with a Sager.... damned thing gets hot enough to warp the plastic it's made of.
  • LeonardoLeonardo Wake up and smell the glaciers Eagle River, Alaska Icrontian
    edited January 2005
    Also, I seem to remember that Intel's published TDP numbers are derated from the actual max TDP of the P4 by around 15%, since they don't think that anyone or any program can keep the absurdly long pipelines in the P4 architecture full and processing 100% on the time. I guess they've never heard of running 2 instances of folding on a HT-enabled P4.

    - Especially since hyperthreading is turned off in the BIOS on probably half the corporate PCs out there anyway! ArrrGHHH.

    - Especially... happy my HT P4 is Northwood and not Prescott. I keep the pipelines full! Yup, not quite as snappy as my AMD rig; but the multitasking is excellent.

    So, what type of cooling advancements, or perhaps true innovations will we see for taming the upcoming dual-core CPUs? Will it be more of the same, that is, same technology, but just more massive? If that is to be the case, it will deal a blow to the trend of smaller desktop computers.
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited January 2005
    RWB wrote:
    Been there with a Sager.... damned thing gets hot enough to warp the plastic it's made of.

    It shouldn't get anything like that hot dude. Either the heatsink is full of crap or something else is wrong.
  • muddocktormuddocktor
    edited January 2005
    Leonardo wrote:
    -So, what type of cooling advancements, or perhaps true innovations will we see for taming the upcoming dual-core CPUs? Will it be more of the same, that is, same technology, but just more massive? If that is to be the case, it will deal a blow to the trend of smaller desktop computers.


    I think that you will see watercooling going mainstream on workstations and high end gaming rigs from the biggest oems like Dell and HP and such. After all, Apple already has done this with their dual 2.5 G5 and it's the only way I can see Dell taming the enormous heat output of the Intel procs fairly quietly. I imagine it will be a tidy all-in-one setup like Apple's setup.

    That's my guess anyways. :D
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited January 2005
    I think you'll see heatpipes first.
Sign In or Register to comment.