XP and heat

cellbiogeek

Distinguished
Dec 31, 2007
230
0
18,680
I've read about the XP's 20% decrease in power consumption but I haven't seen anything about die temperature. Anyone had a chance to see how much cooler (temp wise) the XP is or see any articles about it.
 

Boondock_Saint

Distinguished
Jul 4, 2001
548
0
18,980
I would think so. Core material has not changed, and even so it would not likely affect it.

:tongue: Have you ever tried cooking an egg on your HSF? Tasty. :tongue:
 

Kelledin

Distinguished
Mar 1, 2001
2,183
0
19,780
Not much. I dug up a factoid a while ago--the P4 2.0GHz is actually hotter than any T-bird chip. Heat dissipation is a direct factor of power consumption; the relationship is indeed linear.

I'm not sure about the relationship between power and clockspeed, but from what I've seen, it seems close to linear for the T-birds. T-bird 1GHz CPUs draw ~50W of power, and T-bird 1.4GHz CPUs draw ~73W of power. This means that even with a 12% increase in clockspeed (1400 to 1566), the 1.57GHz Palomino should still run somewhat cooler than the 1.4GHz T-bird.

Kelledin

"/join #hackerz. See the Web. DoS interesting people."
 

charliec2uk

Distinguished
Jul 26, 2001
249
0
18,680
I though that the P4 ran cooler than a 1.4 Ghz T-bird. I think it dissipates 52 Watts. What does the thunderbird dissipate?

Democracy Bernad, it must be stopped!
 
G

Guest

Guest
I read in a review somewhere that the XP 1800 runs almost identical to a t-bird 1.33.
 
G

Guest

Guest
I read in a review somewhere that the XP 1800 runs almost identical tempt as a t-bird 1.33.
 

Kelledin

Distinguished
Mar 1, 2001
2,183
0
19,780
The slower P4's run cooler than a 1.4GHz T-bird, but the fastest P4 (2.0GHz) draws ~75W, according to <A HREF="http://support.intel.com/support/processors/pentium4/thermal.htm#Table2" target="_new">this table</A>. T-bird 1.4's draw ~73W.

Kelledin

"/join #hackerz. See the Web. DoS interesting people."
 

Raystonn

Distinguished
Apr 12, 2001
2,273
0
19,780
A CPU core's surface temperature is a factor of wattage divided by surface area. Because of the Pentium 4's heat spreader, the average surface temperature is actually much less than that of the Athlon. Also because of this heat spreader, there is more surface area in contact with a heat sink, and thus more heat can be dissipated in a smaller amount of time.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
 

lakedude

Distinguished
Dec 31, 2007
1,789
0
19,780
I took physics too and would like to point out the difference between heat and temperature. Lets say you have a big cup of ice. Which will do a better job melting the ice, a small needle at 1000 degrees, or a bucket of hot water at 200 degrees? Clearly the needle is not going to melt much ice because while very high in temperature it does not contain much heat.

Same goes for processers. 75W will raise the temp of a smaller core more then a larger one but the heat will be the same in both.


Remember if you ain't Muslim you ain't Shiite.
 

Kelledin

Distinguished
Mar 1, 2001
2,183
0
19,780
Whoops, a little more data. Apparently the Pentium IV can consume even more power than Intel's foremost datasheet lets on! I remember reading about this in the Inquest articles, but that was...the Inquest articles. It looks like <A HREF="http://www.threecom.de/artikel/amdround1/" target="_new">Inquest was right</A>--the foremost datasheet reports the power draw as only 75% of its maximum. (sorry, babelfish may be necessary for that link) Perhaps suitable for an average...but I believe AMD rates theirs by maximum power draw, not average. Something else to wonder about.

The PDF version of Intel's datasheet notes that the P4 power consumption figures listed aren't actually indicative of worst-case conditions. It tells you to go check the <i>Intel Pentium 4 Processor in the 478-pin Package Thermal Design Guide</i>--which I'm having a difficult time finding. Maybe Google can dig it out of Intel's site...

And from the above site...The AthlonXP PR1800 apparently consumes about 66W, roughly equal to the T-bird 1.2's power consumption. Not bad at all...and the .13u Thoroughbred is supposed to get about a 40% power consumption drop clock-for-clock (compared to the Palomino).

Kelledin

"/join #hackerz. See the Web. DoS interesting people."

UPDATE: Concerning the Thermal Design Guide...

Nope, Google can't find this mystery document. Neither can Intel's site search. Raystonn, can you help?
 

Kelledin

Distinguished
Mar 1, 2001
2,183
0
19,780
<i>The Intel Pentium 4 Processor in the 478-pin Package Thermal Design Guide</i>.

It's referred to by the <A HREF="http://developer.intel.com/design/pentium4/datashts/249887.htm" target="_new">PDF version of the basic Intel P4 datasheet</A>, page 77, Table 32, Footnote 2. A thought comes to mind...is the document in question perhaps a subsection of this datasheet? In any case, I must be blind, cos I'm not finding it...

BTW, It's still spooky how you appear whenever your name is called/typed. :tongue:

Kelledin

"/join #hackerz. See the Web. DoS interesting people."
 

Raystonn

Distinguished
Apr 12, 2001
2,273
0
19,780
You will find that <A HREF="http://developer.intel.com/design/Pentium4/guides" target="_new">here</A>. Note that you should not assume anything about how AMD rates their power requirements (watts.) It is typical that the average high-end rating is given, for both companies. Additionally, equal power requirements for a Pentium 4 and an Athlon would lead to much cooler temperatures for the Pentium 4 due to its heat spreader.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
 

Matisaro

Splendid
Mar 23, 2001
6,737
0
25,780
Unless the p4's heatspreader sucks ass, which may be the case seeing as I read somewhere it was stainless steel. Which is a very poor heatconductor.

(ps: When I read it I asked on a forum for confirmation and never got it, therefore I am not positive the heatspreader actually is steel).

PPS: the heatspreader is just another material which must thermally couple between the core and heatsink, I doubt it actually helps temps, at best it would be the same as not hacing a heatspreader. (extra area versus lower thermal transfer)

^all guesstimates, any hard info to confirm/deny will be accepted gratefully.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
 

Kelledin

Distinguished
Mar 1, 2001
2,183
0
19,780
According to <A HREF="http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/24309.pdf" target="_new">this document,</A> 66W is <b>maximum</b> power consumption of the AthlonXP PR1800. Typical power consumption is ~59W.

Kelledin

"/join #hackerz. See the Web. DoS interesting people."