How hot/cool is your Phenom II

[PoopyPants]

guys i am trying to gather some info about the temps of the new Phenom II.
any of the chips will do but i am more interested in the X3 720 BE than anything else.

if you have one of these chips please post the chip the type of cooling you are using and the overclock and voltages being used.

my X3 720BE is just reporting temps that are impossible so i have no way of gauging the temps.
but if you fellas report back, i can get a good idea of what others see, so i can have a basis for my temps.

i've used the sunbeam Core Contact Cooler (wicked freakin cooler, just wicked!! ) and a new thermaltake model, the CCC is better so temps are even lower! and harder to know whats really what.

 

[Rhoxed]

940BE @ 3.9Ghz 1.525V Xigmatek Dark Knight Idle 40C Load 61C

940BE @ 3.9Ghz 1.525V Xigmatek Thors Hammer Idle 37C Load 55C

940BE @ 4.05Ghz 1.55V Xigmatek Thors Hammer Idle 38C Load 57C

940BE Stock Dark Knight idle 33C Load 48C

940BE Stock Thors Hammer idle 32C Load 46C

also have a 9850BE if you want to know temps on that (from 4 different coolers)

Ambient was 76F on all tests

 

[Flipped Gazelle]

My PhII X3 710 @ 3.4, 1.475v idles at 38C CPU temp and 37C cores, according to Everest.

After 20 minutes OCCT, 51C CPU and 47C cores.

Once I enable the 710's 4th core, temp readings are all borked.

Using a Silenx Ixtreme 120 cooler, fan @ 1000 rpm.

 

[PoopyPants]

yeah i am getting and i apologize for not posting this in the original post (i will add it) 23c idle and 43c load at 1.55v with the thermaltake and about 5deg less with the Sunbeam CCC.
room temp is about 21c

to me thats impossible.
the Sunbeam is a killer heatsink but less than 40c under 100% load at that voltage?? no way.

 

[Summer]

Phenom 720x3 @ 3.6 1.45v w/ Xigmatek Dark Knight: Idle 23c (Core) - 32c (CPU) ~ Load 35c (Core) & 39c (CPU).

Enabling 4th core also screws up temperature reading. System stable at 720x4 @ 3.5 1.45v but 790GX cold boot bug/problem has prevented me from running 720x4 core default. No response from Asrock tech support and it's been a week.

 

[PoopyPants]

Interesting temps summer. 1.45v and on 39c.
granted these are 45nm chips and only a tricore but,,
those are the same temps i am seeing and i dont trust them for 1 minute.

i just cant stand not knowing how hot my cpu is....

 

[Smoove910]

As of this writing, I am running cool-n-quiet enabled on my 720 BE.

IDLE = 25c @ 800mhz, 1.15vcore
LOAD = 52c @ 3.675ghz, 1.52vcore

This is with three instances of ORTHOS going on each core, setting the affinity for each core. Temps reported with Speedfan and Coretemp. Also to note, this is a microATX motherboard in a Coolermaster Elite 340.

I can also unlock the 4th core, but temps are whacked as others have mentioned. I can run all 4 cores @ 3.5ghz 1.475vcore, but temps would climb a little. Not too important to me to have all those core for the added heat, so I shut it off and enabled Cool-n-Quiet.

 

[Summer]

Originally posted by: PoopyPants
Interesting temps summer. 1.45v and on 39c.
granted these are 45nm chips and only a tricore but,,
those are the same temps i am seeing and i dont trust them for 1 minute.

i just cant stand not knowing how hot my cpu is....

I'm also running SpeedFan with it set at 25%-50% gradual climb and full 100% if CPU temperature climbs above 38c. The case also has 4x120mm intake/exhaust fans to move air around. Of course all these numbers should be taken with a grain of salt because accurate temp reading is still dependent on ambient temperature.

 

[Xerran]

At 3.7GHz on my Phenom II 940 I'm idling at 33c and never break 43c at load.

 

[PoopyPants]

well i did more testing last night.
putting a temp probe on 2 sides of the cpu with the thermaltake TMG IA1 i am seeing around 45c full load 1.55v 3.8ghz - thats 1 step extra voltage than i really need but i did it purposely to heat up the cpu.
idle i am showing 27-30c
room temp was 21c if i recall.

granted it is a 95W cpu, but also a 45nm cpu, its curious that temps are that low.
again the probes were on the side of the cpu so one could naturally assume the core was atleast 5c higher, if not 8deg higher.

 

[perdomot]

Got my 720BE going at 3.2Ghz at default voltage and idle temp with a Big Typhoon cooler is 33 with local temp of 27 here in Miami.

 

[PoopyPants]

i always wanted of those monsters. they are showing their age when compared to the naked heatpipe coolers but they are still a great heatsink and appearently, cool the 720BE perfectly fine.

these chips at default or overclocked to xxxghz with default voltage are just cool running mofo's. hell even at 1.55v they are cool compared to my Q9650 which is at 3.85ghz at 1.38V which could never be run on air as temps would be 70c + .

 

[AstroManLuca]

My X3 720 @ 3.2 GHz and 1.35v does around 25C idle and about 35C under load. These are just approximate but I know they're close. I did run it at 3.4 for a while at 1.504v and it was about 10C hotter for both. The extra voltage isn't worth it for just 200 MHz though. My 720 is a dog.

 

[PoopyPants]

wow you shouldnt need that much voltage for 200mhz. unless you got a really nasty chip.
dont forget that you need to increase your NB and hyper transport too sometimes and not the vcore.

but temps you all have given me so far are somewhat on par with mine.
using the probes on each side of the cpu IHS along with my IR sensor i am getting a better feel for the temps.
i figure if my temp probe is showing 40 deg 100% load on the side of the IHS i am atleast 5c higher in the cores... maybe. but i would rather guess higher temps than beleive they are so low all the time.
most programs report core temps in the low 30's, which is just not possible at such a high overclock and high vcore.
they maybe 95w at stock but with an OC they jump very quickly to 150w+ and produce alot of heat if not managed properly.

 

[Accord99]

Originally posted by: PoopyPants
these chips at default or overclocked to xxxghz with default voltage are just cool running mofo's. hell even at 1.55v they are cool compared to my Q9650 which is at 3.85ghz at 1.38V which could never be run on air as temps would be 70c + .

You can't directly compare Intel's temperature readings with AMD's.

 

[konakona]

heh I really have a stubborn chip, looks like.
1.5v @ 3.12 (240 ref), NB at around 2.4ghz @1.4v.
around 32C idle and up to 42C under load.

x3 710 here.

 

[PoopyPants]

Originally posted by: Accord99

Originally posted by: PoopyPants
these chips at default or overclocked to xxxghz with default voltage are just cool running mofo's. hell even at 1.55v they are cool compared to my Q9650 which is at 3.85ghz at 1.38V which could never be run on air as temps would be 70c + .

You can't directly compare Intel's temperature readings with AMD's.

if they are both, for example 65nm and both running 1.4v then the math is similar.
the real world results are not but we all know that AMD chips run cooler.
and yes i know the PII is 45nm.

my point was simple and not meant to be taken as a direct comparison, only that an intel running at lower volts is hotter than an amd at higher volts.
call that a direct comparison if you will, it just shows intel chips run hotter by their very nature.
hell i am an intel guy lol, i know how hot those intels run.

 

[Accord99]

Originally posted by: PoopyPants
if they are both, for example 65nm and both running 1.4v then the math is similar.
the real world results are not but we all know that AMD chips run cooler.
and yes i know the PII is 45nm.

The temperature sensors for Intel and AMD are located in different parts of the die (the PhII only has a single on-die sensor for example) and require different compensation values. There are plenty of reports of the PhII temperatures that are cooler than ambient clearly indicating the absolute accuracy of the PhII sensor is not good.

my point was simple and not meant to be taken as a direct comparison, only that an intel running at lower volts is hotter than an amd at higher volts.
call that a direct comparison if you will, it just shows intel chips run hotter by their very nature.
hell i am an intel guy lol, i know how hot those intels run.

Well based on the excellent power usage of Yorkfields at stock and even at highly overclocked levels, I wouldn't be suprised if 70C on a Yorkfield (measured from its on-die sensors) is cooler than 50C on the PhII (measured from its on-die sensor).

 

[AstroManLuca]

Originally posted by: PoopyPants
wow you shouldnt need that much voltage for 200mhz. unless you got a really nasty chip.
dont forget that you need to increase your NB and hyper transport too sometimes and not the vcore.

Nah, that's after increasing the NB and HT voltage (slightly). Before that I could only get it to 3.3 GHz, and 3.2 is the highest it'll go without pumping the voltage into the 1.5 range. Mind you, it would at least post all the way up to 3.5 GHz... but at 3.4 (before increasing the NB/HT voltage) it failed Prime95 and at 3.5 it BSOD'd at startup (but at least it posted). NB/HT + 1.525vcore at 3.4 was enough for it to be stable, but I felt it wasn't worth it for an extra couple hundred MHz when I could easily do 3.2 at near-stock voltage. So I miss out on 200 MHz in exchange for less power consumption and (hopefully) better component life.

I'm not really unhappy with it, but it's just a counterpoint to those who claim that you can easily OC the 720 to 3.5-3.6 on air. Some can do that. Others can't. Of course, no amount of overclocking is guaranteed.

 

[PoopyPants]

Astro -

i found on mine also that it takes quite a jump in voltage to go from 3.2 - 3.5ghz. not unheard of but somewhat surprising to see such a jump needed.
then it sort of levels off, then i need to increase it more.
typical overclocking walls we see in most chips.

 

[Summer]

Turning into another Intel vs AMD thread. Typical internet e-penis contest.

 

[PoopyPants]

wow... typical response from someone who doesnt have a clue nor read one single word of anything that was said....
yeah... you should stick to the noobie section

 

[Malakai]

Mine also requires a lot of volts for 3.2->3.5.

At 3.2ghz it runs great at 1.29v. To step up to 3.4 requires 1.4v and 3.6 requires 1.475v.

My HSF is not the best, old SI-97A. Im going to put a sunbeam core contact on it, at least then I will be comfortable at 3.6ghz without temps going over 60c on load.

 

[Rockinacoustic]

Using the stock cooler, I bumped the multiplier up to 16x and have it running stable at 3.2GHz on stock voltage (1.28)

20min of OCCT gave me a high temp of ~45*C, and with Cool n' Quiet it idles around 23*C at 800MHz