PDA

View Full Version : are donthans as cool as we think they are?



FireDragon
12-17-2005, 01:08 PM
at lower speeds i would say yes but then agin i was reading a thread in the air cooling that a dothan at 2880 at 1.6 wa at 47 c with a 120...so i mean how can that be cool when a fx-57 at 2.8 wont reach those temps?!?! just wondering...

Dragon

Elisha
12-17-2005, 01:40 PM
but a Dottie at 2.8 will smoke an FX57 @ 2.8 anyday :-)

alexio
12-17-2005, 01:43 PM
at lower speeds i would say yes but then agin i was reading a thread in the air cooling that a dothan at 2880 at 1.6 wa at 47 c with a 120...so i mean how can that be cool when a fx-57 at 2.8 wont reach those temps?!?! just wondering...

Dragon

1.) The conditions probably weren't the same.

2.) The Dothan core is much smaller.

3.) Temperature messurement is often not very accurate.

mrlobber
12-17-2005, 01:46 PM
Then again, I'm using a Dothan 740 at 2.6 Ghz passively cooled daily ;)

[cTx]Philosophy
12-17-2005, 01:55 PM
yah Im AMD fanboi 4ever
Buttt... Ive been thinking of bulding a dothan rig next just because of pure performance of it 3d wise..

NiCKE^
12-17-2005, 01:56 PM
Someone said that the temp reading on dothans are wrong, I believe it's with very big numbers they are wrong too. Something with the adapter which ***** things up.

FireDragon
12-17-2005, 01:58 PM
just a question that i had...true it is smaller die size same 90nm though...smoke it no beat it yes games yes in benchs yes in just about ANYTHING else umm...no...lol any way i was just wondering i was actually looking into a dothan for a while then decided to stay with my A64's i dont hant intel i just prefer AMD to it that is all...so ya any way i was just wondering as that seemed REALLY hot for a chip that is supposed to only put out about 65watts at around 3ghz...even if the core is half the size it is half the heat...as the fx...

Dragon

FireDragon
12-17-2005, 02:00 PM
Someone said that the temp reading on dothans are wrong, I believe it's with very big numbers they are wrong too. Something with the adapter which ***** things up.


well i mean how wrong can it be that does seem high then agin in one way it seems kinda right...

Dragon

saaya
12-17-2005, 03:02 PM
its the epentium m termal probe i think, its reading much higher temps than other cpus.

its always problematic to comment on this, as inside a cpu you have several different heat zones. some parts run hotter than others. some parts inside a cpu can get as hot as 100°C or more, but its actually ok and wont hurt the cpu.

the cpu temps we are used to see are not the max cpu temps that the hottest part of a cpu runs at!
the dothan temp probe is probably more close to what the max inside temp of the hottest part inside the cpu is, but we cant compare this to other cpus termal readings.

and i think its not the ct479 or the bios of dekstop boards that has to do with this, its the termal probe. on centrino laptops the cpu temps are also very high i heard, i think its the way the termal probe gets read and where its placed inside the cpu.

but i do think that the desktop mainboard bioses read the numbers even higher than they are actually. so the pentium m cpus report a higher temp AND the mainboards report those temps yet again even higher.

ive heard of up to 110°C under load without any problems, and i had it up to 85°C under load with a 780es at stock speeds and stock vcore!
and the heatsink was ice cold... so dont take the temp readings seriously :D

EDIT: usually there seems to be an unofficial agreement to adjust the termal probes in a way that the temperatures they report are what the temperatures on the surface of the cpu or the base of the heatsink/ihs are. more or less... thats at least the expirience i made when meassuring cpu temps with a probe. on pentium m chips this is different.

think of gpus as well for example, they report temps way higher than cpus as well, my 6600gt is supposedly running at 45°C in idle while the heatsink is totally cold to the touch...

wwwww
12-17-2005, 06:20 PM
Yup, my lappy is at 61C atm idle @ 2GHz 1.34V - doesn't even bother turning the fan on, heatsink is around 35C though. Bit of load, the heatsink goes to around 50C and the probe up to around 70C.

Although, on the DFI 855 I found temps ram alot lower - According to the probe I had it idle below ambient, and running heatsinkless, I had it load at 51C but already was too hot to touch.

saaya
12-17-2005, 06:46 PM
i think dfi adjusted the temp readings on that board to be comparable with other cpu temps. thats why the temps were lower i think :)

lowfat
12-17-2005, 07:03 PM
I think they do run quite cool. I have my CNPS700B-AlCu running @ 5V, 2650MHz, 1.45V, and the HSF is never warm to the touch.

fordf250
12-17-2005, 08:06 PM
I think that heatpipe coolers need the stuff inside to be liquid at room temps or above or the change in state would not work,so the temps will not get much lower than what it changes at under load.44-47C is were my si-97, xp120 and dual xp-120 run with the fans on low or high with a load.

ingentingmendeg
12-17-2005, 08:43 PM
they have to run cool if my EE runs at 39c idle :P

cirthix
12-17-2005, 08:50 PM
55w (2.6ghz, 1.6v, most oc'd dothans here run about 55w) of heat is still a lot of heat to deal with. at .2degc/w (the effeciency of a really good air cooled setup), that's still 12c above ambient. that's assuming that the die is the size of the bottom of the heatsink. if you take a look at how small the dothan die is, you'll understand why 15-20c above ambient is normal.

http://www.benchtest.com/calc.html

real:7700cu: 55w, ambient of 28c, load of 50c with reduced fan speed and a heatspreader
thoretical (with huge die): 55w, ambient of 28c, 39c load with max fan speed

the thing is, the dies are tiny, 55w is a substantial amount of heat, but its not much compared to the 100+w of heat that a comparable cpu would use. in comparison, they are very low power, but on their own, 55w is imo, a lot of power

Daytime Dreamer
12-17-2005, 11:07 PM
My Dothan folds 24/7 for the last 2 weeks and its at 2555mhz with a P4 heatspreader on top (thanks Unseen!) lapped good, some arctic silver 3 and the stock Ct-479 cooler (lapped) and the heatsink is always ice cold although the BIOS says its at 40-42 Celcius :rolleyes:

This is all with 1,425 volts in Bios 1,402 actual

lawrywild
12-18-2005, 04:24 AM
yes, they do run cool, and a dothan @ 2.8 does smoke an fx-57 @ 2.8, it will beat a fx-57 @ 3.0-3.1 (but not smoke it :p: )

anyway, anyone else noticed that in the p4c800-e bios, it reads temps higher than in asus probe (or mbm5 or whatever), bios can sometimes say 42'C and probe usually says around 32'C... :rolleyes:

Highland3r
12-18-2005, 04:52 AM
Could explain why the Dothan reads as 45 deg and the X800's at 17 when running w/iced water... The CPU block is cold to the touch too, so it follows that the sensor could be well off.
Ran a celly M (1.4ghz) on stock sink with fan off for 50 hours prime stable, likewise 730 @ 2.5ghz fan at 50% for 50 hours... On both occasions temps never went above 55 deg... Thats pretty impressive for stock sinks and silence...

Despite "smoking* an FX in games and similar, would still rather run A64 as a 24/7 rig... They're not as fast for games, but general useage win hands down..
(sorry for OT btw)

AgonxOC
12-18-2005, 10:21 PM
Mine is at 2730Mhz @ 1.5V and it never goes past 44º Celcius. I use a Zalman 7000CU Led..

Alex

saaya
12-19-2005, 11:39 AM
are those the temps reported by the board?
interesiting, looks like the bios is adjusted to read the temp "correctly" on that board ...

AgonxOC
12-19-2005, 12:14 PM
Thats what MBM5 reads to me... 28ºC at idle (hot summer day) and 44ºC max in the case outside the case it never went above 41ºC.

Alex

saaya
12-19-2005, 12:28 PM
that sounds like they adjusted the probe reading, temps sound like they should be comparing to other cpus

crotale
12-19-2005, 02:17 PM
anyway, anyone else noticed that in the p4c800-e bios, it reads temps higher than in asus probe (or mbm5 or whatever), bios can sometimes say 42'C and probe usually says around 32'C... :rolleyes:
That is quite comon for a lot of boards, and one reason is that the BIOS stresses the CPU to some degree...
Not really stressing it, but it's not idling it like windows does.

saaya
12-19-2005, 03:36 PM
i think that asus didnt adjust the readings in bios but only in windows.
somehow the adjusting of the temp probe seems to only work in windows i guess...

lawrywild
12-19-2005, 04:06 PM
im pretty sure my dothan reads temps corrently in probe but not bios..

I'd thought of the idea that asus had adjusted it or something a while back, and it still seems thats probably what they did

@crotale: but the bios reads temps higher than load in windows, so it can't be right..

[cTx]Philosophy
12-19-2005, 05:03 PM
whats best mobo to pair up with a 770?
and cooling etc
I want a 533 cpu and air cooling

speed bump
12-19-2005, 10:08 PM
Best mobo would be a P4c800 or a P4gd1 if you got luckier than any person should and found one.

Cooling I would go with something that alot of people have had fit on there mobo but I am not sure exactly what is the best

crotale
12-20-2005, 12:30 AM
im pretty sure my dothan reads temps corrently in probe but not bios..

I'd thought of the idea that asus had adjusted it or something a while back, and it still seems thats probably what they did

@crotale: but the bios reads temps higher than load in windows, so it can't be right..
Well, that sounds strange though.

Personally, I don't care much about the temp read-outs.
In my own testings, the Dothan put out way less heat than a FX
FX @ 3GHz 1.65vcore and I get -40C on my single stage
730 @ 3GHz 1.65vcore = -50C