CPU Heat Comparison: How Hot is Prescott?

by Derek Wilson on 4/16/2004 4:32 AM EST
Comments Locked

48 Comments

Back to Article

  • Xentropy - Monday, April 19, 2004 - link

    "Most of us here are not mindless zealots who believe everything we read."

    Unless it's anti-Intel propaganda.

    "but i have seen a prescot melt motherboards"

    Have you? How many motherboards? Online in some forum or on HardOCP or whatever doesn't count. Personally seen, in person, how many?
  • Pumpkinierre - Sunday, April 18, 2004 - link

    #46 I thought that was standard with Win XP. When I looked at the idle temps., I thought that might be the problem. On my K6-3/Win98, I run Rain which does the same thing. Good Post!
  • lesovers - Sunday, April 18, 2004 - link

    The main reason the AMD XP chips are so much hotter at ilde is they do not normally implement the HALT command like the P4s. The VIA KT chipsets (266, 333 and 400) have registers that can enable the CPU to HALT. Link for this is:

    http://www.ocmodshop.com/default.aspx?a=125%20

    My XP2000 cool down be by about 20C at idle with the chipset registers changed !!!!!!
  • boardsportsrule - Saturday, April 17, 2004 - link

    ok wow, how can you have a 3.2 prescot be at 34C with a medeocer heatsink, when i see people complain about high temps with their 2.8 prescott using a sp-94...then onto how could a 2600+ be hotter then a 3.2 prescott...i havent seen a 2600+ melt a motherboard, but i have seen a prescot melt motherboards... this stinks of intel...wounder howmuch they are paying him?
  • ZobarStyl - Saturday, April 17, 2004 - link

    This processor is supposed to carry Intel through something like 4.0+ GHz when Tejas gets here; if it's this hot against a Northwood, how do they expect to hit past 3.8 or so without watercooling or Prometia's built into standard desktops. The question isn't whether or not Prescott is a bad processor (it's not a disaster of Williamette proportions) but whether or not it can carry the weight Intel is slated to heft onto it's back for the next cycle.
  • mindless1 - Saturday, April 17, 2004 - link

    Margalus, get a grip!
    Even INTEL specs it to run that much hotter. Will you not even believe Intel?
  • mechBgon - Saturday, April 17, 2004 - link

    Margalus, I've seen several Forum members try the Prescott and go back to their Northwoods because the Prescott was running very, very hot and/or doing scary stuff to their motherboards. And these are neither AMD zealots nor newbies. IIRC one of them is running a Prometia phase-change cooling system and another is running watercooling.
  • Margalus - Saturday, April 17, 2004 - link

    I can't believe the number of amd fanatics that just can't accept reality. the prescott is not as hot as the anti intel fanatics are saying. This article was pretty well done for the comparisons they did. All of you saying how it is bs have never even used a prescott, you just listen to clueless people like cramitpal who haven't even passed 3rd grade yet
  • AIWGuru - Saturday, April 17, 2004 - link

    "Of course, we'll be there to test it out as soon as we can get ahold of a 64bit enabled chip."

    Why are you waiting for a 64 bit prescott? Why not use the A64 you already have?
  • AIWGuru - Saturday, April 17, 2004 - link

    Coruscant, your understanding is incorrect.
    What you're TRYING to describe is the flow of (thermal) energy to the path with the least resistance - the movement of hot to cold.
    What you forgot to take into account is that once the heatsink starts to "absorb" heat it may quickly cease to become the path of least resistance.

    Regardless, this test is fundamentally USELESS as, unless the dissipation rate of the heatsink has been exceeded, any measurement will be the same.
    Because heat does not build up on a linear scale, any measurement above that level will also be innacruate.
    The placement of the probe is also somewhat stupid.
  • Etacovda - Saturday, April 17, 2004 - link

    Yeah, id have to agree that the test was next to useless. The methods used were... archaic? to say the least. Put it this way, the AXP thats moved 1 deg under load generally should move in the region of 10 at least. so 1/10, applied to the prescott, would be a lot more... i know this method in itself is flawed ;) but thats the ghist of whats going on here.

    I hope you understand my point ;p
  • Pumpkinierre - Saturday, April 17, 2004 - link

    #35 the heat capacity of air is small. If no heat sink were applied, MORE heat would be lost to the underside/mobo than the topside/heatspreader because the pins/traces/mobo act as a heatsink. Sure, if you place a liquid Helium 'cold finger' on the cpu you will get less heat loss from the underside but remember those metallic pins are linked to the heart of the die so thermal loss is inevitable. Slot 1/A would be better to cool (either side) but there would still be pin connection loss. If you replaced the die in the cpu packaging by a die size heating element with known power characteristics you might be able to produce a calibration curve in conjunction with fluid cooling or a temp. probe that would then be used to determine cpu power consumption. Of course, with the heating element measurements, the mobo is switched off with the cpu package/heating element sitting in the mobo sckt. I suspect Intel/AMD use a power flow determination as direct measurement is difficult.
  • rwong48 - Saturday, April 17, 2004 - link

    like someone else said, i was wondering on the accuracy of these temperatures because the xp 2500+ barton only increased 1ºC from idle to load..
  • Coruscant - Friday, April 16, 2004 - link

    If you provide a cold enough heat sink (not the typical heatsink, but a source that absorbs heat) then the heat transfer to that heat sink will override the transfer from the processor sides/bottom to the ambient air. Additionally, so long as the ambient conditions are maintained, it would be reasonable to assume that the heat loss to through those surfaces would be comparable. Of course, assuming the surface temperature of all the processors in question are maintained at a pre-defined setpoint, the heat transfer from the processors to ambient would largely be the same, varying only by the difference in exposed surface area. Again, this would all be in the pursuit of determining at what rate of heat is generated be the various processors.
  • MoronBasher - Friday, April 16, 2004 - link

    *unless i am wrong and they did enable cnq

    but the review wasn't that good altogether.
  • MoronBasher - Friday, April 16, 2004 - link

    the article was missing something. they should have used the diode along with the probe. also, they should have used cool n' quiet for the athlon64 as that would be what is actually going on in real world terms.

    just my 2 cents
  • mechBgon - Friday, April 16, 2004 - link

    The "percent increase" graph should be using Kelvins as a basis, not Celcius.

    For example, let's say a processor idles at 40°C and rises to 60°C under load. That's an increase of 20 Kelvins, or 20°C if you prefer, but it isn't a 50% increase... 40°C is 313 Kelvins, so a 20° rise is 20/313 or 6.4%.
  • AIWGuru - Friday, April 16, 2004 - link

    The methodology used in this "test" wouldn't pass muster in a grade 4 science class.
    I can't believe that ANYONE is THIS inept.
    This article leaves me thinking that someone at Intel paid someone to write it.
  • ZobarStyl - Friday, April 16, 2004 - link

    p.s. Not that I own one of those hunks of junk, but let's be honest that's where intel gets its cash, not from the enthusiast market.
  • ZobarStyl - Friday, April 16, 2004 - link

    I would like to see a closed-case standard (same case and components, diff proc/mobo) with the internal case ambient being measured in different places. If Prescott is such a current hog, then (as was mentioned in relationship to electrical resistance above) other components such as the PSU might start adding heat to the total system. I don't run all these 80mm's to keep just the processor cool; I run them to keep all my components cool and keep system life at an optimum. If Prescott under load is 50 degrees but the sum of its output makes my case 15 Celsius hotter, then no amount of CPU speed will make me want to buy that. A processor is just one part of the system, albeit the most important, but that doesn't mean I want to buy every other component in my system twice because of premature thermal breakdown.

    Every computer case is a thermodynamic system straight out of a Physics textbook. Just because one component makes 10 degrees more heat doesn't necessarily mean the whole system can dissipate all that heat, and BTX aside Intel is going to have some serious issues running these things in major manufacturer style cases. I mean when the best cooling setup you get with a Dell is a single 80mm, and their smaller ones have no case fans at all, the Prescott is going spell early death for a lot of these systems.
  • TrogdorJW - Friday, April 16, 2004 - link

    AMDscooter, regardless of the fact that you might run a P4 rig, any accusations such as yours ("Speaking of stoopid. Intel must have floated him a good size check to write this article saying the Prescott is not much hotter than the Northwood. The whole article reeks of Intel PR hard at work...") are hard to view as anything but you just being an ass. Do you have any real evidence that Anandtech is getting paid by Intel to provide higher scores? No. Must be pretty comfy taking pot shots from the cheap seats, though.

    There's a major difference between pointing out the flaws in an article in a polite manner (i.e. I don't think your testing methodology is good because....) and just flaming away like a fourteen-year old that needs more Ritalin. Is the article flawed? Undoubtably. But short of spending a shload of time and doing enough work for a master's thesis in thermodynamics, you wouldn't expect perfect results.

    By no means are the results given in the article portrayed as accurate. The entire tone of the article is basically one of, "Well, this is what we did and these are our results, but we're not really sure if they're good or not." Clearly, the response is that the results aren't exactly meaningful. Fine. Now try and be helpful in how to correct the tests (like PriceGaz, for instance) instead of being rude.

    You act as though you're paying AT money for the articles they post. I figure the amount of ad revenue generated by the visits of one person in a month amounts to a few cents at most. Please try to make your two cents' opinion count in the future, instead of squandering it on insults.

    For the record, I also read THG regularly, along with Ars Technica, HardOCP, Firing Squad, and various other sites. I would wager that *none* of them receive money from Intel or AMD in order to get a better review. Money for ad space? Sure. Free product to review? Probably. But I've seen every one of these sites rip on both Intel and AMD products at various times, so just get off your high-horse and contribute, people.
  • Pumpkinierre - Friday, April 16, 2004 - link

    If you use a coolant, you still only get the heat emitted from the top half of the cpu unless you had a fully enclosed 'Powerleap' type water bath adaptor- difficult. The only other way is to measure power in and power out on the all the pins. Possible but difficult and it doesnt allow for calculation energy consumption but this I suspect would be minute cf to heat production in a cpu. These methods also dont identify heat localisation in a cpu which is just as important and limiting. In the case of Prescott with its 90nm, switched off 64bit and large L2, I would suspect this is a real problem ie hot and localised. So even though derek's methodology is crude, its still a reasonable stab at a difficult problem.
  • mindless1 - Friday, April 16, 2004 - link

    It was disappointing that bus disconnect wasn't enabled on the Athlon, so it was quite a dissimilar comparision at IDLE.

    The article has one conclusion quite backwards-

    "In other words, since Northwood is cooler than Prescott, our thermistor might be getting cooled even more by the fan. This could mean that Prescott and Northwood are even closer in total heat dissipation than in our temperature measurements."

    If the Prescott is heating up the heatsink more, there's a greater difference between it and ambient, the same fan, flow-rate, is more effectively cooling the hotter CPU. A 1C temp difference in the heatsink corresponds to a larger difference in the CPU.
  • jojo4u - Friday, April 16, 2004 - link

    I'am disappointed that you did not enable the disconnect on the Athlon XP and Duron platform. That would place the Ahtlons much better. I am also interested wether Cool'n'Quiet has been enabled.
  • eetnoyer - Friday, April 16, 2004 - link

    I think a good way to approximate thermal dissipation would be to have a normal water cooling setup with everything insulated except the radiator. The only addition would be a flow diverter to swith between the radiator and a bypass. When the system reaches temperature equilibrium (most easily measured by a simple thermal probe in the reservoir) swith from the radiator to the bypass. Measure the temperature over a given amount of time or until a specific temperature is reached. You could easily compare processors against one another. Or, if you know the precise amount of water in the system (not including the radiator loop), you could calculate heat output for each processor.
  • Phiro - Friday, April 16, 2004 - link

    Anandtech: Good article, you did your best to sort through a apples to oranges to bananas to grapes type of situation.

    From the way people talked I expected huge heat differences and this has altered my outlook on Prescott abit. Thanks again.
  • tyski - Friday, April 16, 2004 - link

    I would also like to see some sort of measurement involving the mass flow rate across the processor and a the temperature at inlet and outlet. Having also suffered through thermo hell, Derek's method reeks of uncontrolled outside factors, such as the case temp affecting the cooling rate of the fan and the heat transfer of the heatsink to the outside air. Temperature != power and power is what I think everybody here is interested in.
  • theobscure - Friday, April 16, 2004 - link

    a better way to test the heat generated by the processor would be to construct walls on all 4 sides of the processor, and then fill them with a known amount of water, then run the processor for a set amount of time and measure the change in temperature of the wath bath.
  • CruisinGT - Friday, April 16, 2004 - link

    In regards to the author's last statement, I would love to see a Prescott and motherboard compatibility article particularly on the first i875 motherboards released like the AOpen AX4C Max and the original Abit IC7s!! Thanks.
  • AnonymouseUser - Friday, April 16, 2004 - link

    "Boy, some of you guys are cruel assholes and belong in Tomshardware and not in anandtech with your attitudes."

    Uh, not! If they did belong at THG, they would have eaten this up as "The Gospel". Generally speaking, Anandtech readers can spot bullsh!t better than THG readers, and I have to agree that this article seems to have a whole lot of "Intel Inside"...
  • AMDScooter - Friday, April 16, 2004 - link

    Quote:"Boy, some of you guys are cruel assholes and belong in Tomshardware and not in anandtech with your attitudes."


    Most of us here are not mindless zealots who believe everything we read. We post what we feel are valid rebuttals to an article, so we are asshoels now..?? Whatever. At the very LEAST the article is a bit misleading. I can see no other point to the article other than getting uninformed peeps like Icewind to buy Prescotts... My2c.

    BTW, dont mistake my name for bias. I currently run a 3.2c in a Vapo unit and am very happy with it.
  • steller2k - Friday, April 16, 2004 - link

    Having some background in thermodynamics, electrical theory and computers (The Navy slang term was Nuc Electrician). It occurs to me that the Prescott processor itself may not be much hotter than the Northwood. The lower voltages would lower the losses in the die reducing the amount of heat there, offset by the increased current requirements and increased losses due to the closer proximity of the transitors, hence roughly the same performance. The biggest problem as I understand it is the motherboards themselves overheating. The term is I^2R losses, as current increases the actual heat generated in other power supply and motherboard components are squared, or in other words if current requirements are doubled the heat produced is four times the original amounts. Don't get me wrong, I really like Anandtech and I found this article surprising and informative; I didn't expect the temperatures on the Prescott and Northwood to be that similar. But I think in the future that for a clearer picture of total heat cost of a processor should include measurement of the the ambient case temperatures. An interesting side note: the higher temperatures of the P4s helps them to dissapate more heat than the lower temperatures of the Athlons, of course that only works up to the point that the rest of the computer doesn't melt. ;) I'm trying to think of a few ways to measure the true heat cost of a certain set of components, but most of what I'm thinking would require some significant mods or a custom/purpose built case. (Like the heater used for heatsink testing mentioned above).

    Relatively happy owner of an AMD Athlon 64 FX, if only there were some worthwhile apps and Socket939 had been a launch item... <sigh>
  • AnonymousPutterer - Friday, April 16, 2004 - link

    Crouscant, you might like to see some notes on my rather sloppy DIY rig, featuring precisely measured flow rates and temps before and after:
    http://www.crystalfontz.com/forum/showthread.php?s...
  • TauCeti - Friday, April 16, 2004 - link

    I just don't get what the author of this article wanted to measure. Was it the temperature of the die in combination with a specific cooler?

    If it was about _heat output_, why not just measuring the power 'usage' of the CPU? I think about 100% of the 'used' electrical energy is converted to thermal heat _inside_ the case because it does not look like the CPU radiates HF-waves and induces heat _outside_ the case?!

    I just don't get it.
  • klah - Friday, April 16, 2004 - link

    Add another graph showing the idle temps of the A64 with C'n'Q enabled.

  • Icewind - Friday, April 16, 2004 - link

    Boy, some of you guys are cruel assholes and belong in Tomshardware and not in anandtech with your attitudes.
  • Sahrin - Friday, April 16, 2004 - link

    Pardon my lack of a degree in thermodynamics, but isn't there a significant difference between heat dissipated and temperature? We were shown that the effectiveness of heat distribution negatively impacted the XP's quite significantly (and this is further supported by the fact that they didn't increase percentage-wise in temperature anywhere near as much as the heatspreader processors did)...aren't there possibly other factors here? Intel said the TDP for Northwood was...what, 89 as I recally? And TDP (this is an average, not a maximum) for the Prescott is 103W, and it is speculated that Prescott at times will dissipate as much as 130W? Isn't this what is important...? (I may just be running in circles here) The amount of heat put out into the volume of the case...the heat doesn't just disappear, it fills the volume of the case and then is blown out (or ideally, is blown out before it has a chance to dissipate within the case)...this is the reason for the BTX standard, new huge coolers the LGA etc. etc. etc. Right?
  • AMDScooter - Friday, April 16, 2004 - link

    Speaking of stoopid. Intel must have floated him a good size check to write this article saying the Prescott is not much hotter than the Northwood. The whole article reeks of Intel PR hard at work. WTF kind of testing methodology is his temp probe in a heatsink BS?? All the chips he tested have an on die temp sensor, but he takes all his readings from a hole drilled in the heatsink?? What a load of bunk. I thought Anand had finally grown up as he has not written an article this biased in some time. For a few minutes there I thought I had accidently stumbled onto Tom's site. So at this point our choices are to believe the rest of the world(myself included) and forum members all over in saying the Prescott is hot as h3ll or.... believe Anand and run out and buy a cool running 3.2E.... and watch it melt our MOBO's.
  • Jeff7181 - Friday, April 16, 2004 - link

    Whew... glad... I thought I might have been the only one who read this and thought "huh?"

    This doesn't appear to be your best work, Derek.
  • Coruscant - Friday, April 16, 2004 - link

    Maybe some suggestions on how to better measure the heat load of the processors. Providing a mass flowrate of coolant across the processor sufficient to cool the processors to a set temperature, and with the coolant having a known inlet temperature condition, measuring the outlet temperature would provide a reasonably accurate measurement of the heat generated. Considering how commonplace liquid cooling has become, certainly this setup wouldn't be out of the question. All that would be required would be a refrigeration loop, adequately insulated coolant reservoir, variable speed pump, and several (3) thermistors. Just a suggestion, from a mechanical/process engineer.
  • PrinceGaz - Friday, April 16, 2004 - link

    Interesting article, but I've a couple of suggestions :)

    If the fan on the heatsink was plugged into the motherboard used to test the CPU, you should instead run it off a dedicated stable regulated power source given an accurate known voltage. That totally isolates any effect the motherhoard, PSU load or anything else might have on HSF efficiency.

    All current and future CPUs we're interested in have a heatspreader, which presumably does a decent job of spreading the heat over a fairly wide surface area compared to core-size of the processor (the AthlonXP/Duron line is effectively finished and won't concern enthusiasts, so their energy-consumption isn't too important). It seems a good way to translate recorded temperature over ambient temperature, into an actual power consumption in watts would be to make a little heater unit (with something like a heatspreader) you can mount the HSF on.

    By adjusting the voltage applied to it and measuring the current you could take a whole range of readings covering a wide range of watts. You'd not only know whether the temperature difference rises linear with power, but also be able to give actual wattage figures for the CPUs. At least one site (whose name evades me) which does extensive HSF testing actually uses such a heater type system to compare HSF performance to ensure identical conditions.

    From the results I must say I'm impressed with how well the Athlon 64 CPUs did, especially when running Prime95. Was Cool 'n Quiet enabled to reduce their speed in idle mode? If it was, I assume you did ensure with some utility they were operating at full speed while Prime95 was running? ;)

    Prime95 is a great way to heat up processors but it may have a different impact on different CPU designs. The P4 runs it *FAR* faster than the AthlonXP does, or even the Athlon 64 despite the latter also having SSE2 instructions (which Prime95 uses). In particular have a look at the benchmarks page:

    http://www.mersenne.org/bench.htm

    and you'll see that for equivalent tests, the time taken by some different CPUs were:

    Athlon XP 3200+ - 0.032, 0.066, 0.142
    Athlon 64 3400+ - 0.025, 0.052, 0.116
    Northwood 3.2GHz - 0.013, 0.026, 0.056

    The Northwood 3.2GHz is running Prime95 roughly twice as fast as even the Athlon 64 3400+. Now while I can't say for sure that running it faster translates into more power-consumption, the fact that the test runs so much faster on that design makes the results difficult to trust. If however you've tried a whole bunch of different stress-test programs on the Athlon 64, Northwood and Prescott and found Prime95 gave the hottest temperatures on them all, thats fine. But if different tests give the hottest temperature for each design, then thats the one you should use for that design when carrying out future tests.

    Sorry for rambling on so long, hope some of that made sense :)
  • CRAMITPAL - Friday, April 16, 2004 - link

    http://www.theinquirer.net/?article=15374
  • CRAMITPAL - Friday, April 16, 2004 - link

    Obviously no one with a clue would buy a PressCrap when it clearly runs much hotter than the Northwood and is actually SLOWER in system performance. In InHell marketing speak that's called "90 nano performance". Sic !!! Now inHell is doing damage control after lower than expected profits and defective products they can't even deliver in quantity to Dull and WasteGate.
  • Mumrik - Friday, April 16, 2004 - link

    I meant an articles that claim the AXP only has a difference of 1
  • Mumrik - Friday, April 16, 2004 - link

    Doesn't seem very well done....

    How can we trust results that claim the AXP2500+ only have an idle/load temperature difference of 1 degree celcius?!
  • Pumpkinierre - Friday, April 16, 2004 - link

    You didnt include the ambient air temp close to the fan (and fan speed- maximum or silent). Northwood temp range seems high. Cpu case temperature can be 10C under internal/diode temp.. 5C or so difference, both idle and load, between N'wood and Prescott is quite significant to an overclocker. The die size is approx. the same but given the 'missing 30 million transistors', 90nm and the large cache, the heat might be more localised on the Prescott affecting headroom. Good to see the a64 data. These AMD cpus seem to run cool.
  • Voodoo80 - Friday, April 16, 2004 - link

    What about cool and quiet? It's a pity this is not tested. I want to know what the idle temps would be if this is enabled.
  • MAME - Friday, April 16, 2004 - link

    Awww, I was hoping it'd be a heat nightmare and AMD would come to save the day!

Log in

Don't have an account? Sign up now