Bonjour et Bonne année à tous !

Je viens d'upgrader mon MacPro 2009 4.1- 2x 2,26 GHz en 5.1 @ 2x 3,33 Ghz .



Avant l'upgrade j'avais les températures suivantes :

En utilisation normale (internet par exemple) j'avais: CPU A : 38 °C et CPU B 25 °C .

En utilisation intensive(Pro Tools avec 30 instances de kontakt par exemple) j'avais CPU A 52 °C et CPU B 43°C.


Après l'upgrade avec deux XEON 5680 j'ai les températures suivantes :


Sans rien faire tourner (aucune application) : CPU A 54°C et CPU B 45 °C .

En utilisation intensive : CPU A: 87°C (voir plus !) CPU B : 65°C

Voilà je voulais savoir si ces températures sont normales ou s'il fallait que je m'inquiète.

PS : J'ai démonter puis remonter les processeurs plusieurs fois en veillant à chaque fois à bien appliquer la juste quantité de pâte thermique(Artic),Intel recommande une ligne droite d'1 millimètre sur la longueur du CPU pour les XEONS X5680 (sans idle) et appliquer la même pression sur les vis du radiateur qu'il y avait sur mes anciens processeurs c'est dire 7 tours de clé.

 

Last edited: Jan 4, 2017

its no secret that macbook pros get hot, especially with your upgraded cpu.

these temperatures are too high, but not fatal over short periods of time.

try to go desktop for big projects and install proper cooling.

-whats your point exactly, this is a little too much French for me ;)

 

Thanks for your help

However it's not a macbook but a MacPro (desktop)

 

It's obvious that your new chip runs much hotter where it shouldn't.

About applying the paste... I am not sure that's the way to go. If you put a straight line from edge to edge, isn't that going to spill over edge when you press with the cooler? o.O

My hottest CPU is 5820k, runs OCed at 4.4 with temp around 50C full load (3 PCs here). This is what I do... first he explains the wrong ways:

But i'm not sure that wrong paste applying can make such a difference... maybe it can, no idea really.

 

Yes, it can make a bit of a difference, depending on few factors. Silicone-based thermal compounds should always be applied using "peas" method. As for which paste to get, I'd go for Shin-Etsu X23-7782. It's very good for its price.
Gelid GC Extreme is a bit better, but a lot more pricier.
Finally, Liquid Ultra is the best, but requires special care during application and not compatible with aluminum surfaces.

 

Thanks for you help !

Its the same CPU (IHS removed) like this guy on the following video :

 

Thanks Andrew for your help

I have used this product: " Arctic Silver 5 Thermal Compound 3.5 Grams with ArctiClean 60 ML Kit"


What is the "peas" method Andrew ?

This is the 2 CPU I bought :2x Intel XEON 5680
http://www.ebay.fr/itm/361753825971?_trksid=p2057872.m2749.l2649&ssPageName=STRK:MEBIDX:IT

 

Last edited: Jan 6, 2017

It's about making a pea-size blog in the centre of the CPU heat-spreader and then push on the heatsink so it evenly spreads across the chip:

AS5 is a good paste too.
Looking at the picture now, do your CPUs have a metal heat spreader on them like this:

or are they bare chips as seen on your picture?

 

Tu devrais rechercher quelles sont les températures optimales d'utilisation que préconise intel pour ton cpu et fais en fonction, mais du peu que je sais plus un cpu est froid mieux c'est, sur mon Mpro 4.1 j'ai un xeon w3560 qui oscille entre 30/40 degrés l'hiver et 50/70 degrés l'été, il est de 2008 pour le moment il n'a pas claqué, pour compensé j'ai mis des ventilos branchés sur des transfos réglables ce qui me permet de gérer la chauffe.

PS: Il est entièrement d'origine pâte thermique comprise, presque dix piges comme quoi... Et je suis sous win7

 

Thanks Andrew !

No, they don't have metal heat spreader .


I have the same MacPro and did the same upgrade as this guy on the video :
"2009 Mac Pro 12 Core upgrade (2x Intel x5680)"

 

Last edited: Jan 4, 2017

Ils ne préconisent aucune températures optimales mais indiquent clairement la limite à ne pas dépasser c'est à dire: Max 75 °C .

 

http://www.arcticsilver.com/intel_application_method.html#

So this is what I did (QP3)=

 

Last edited: Jan 4, 2017

Hi friend, you have to put just the equivalent of a grain of rice no more on the processor
;-)

ps: Bien nettoyer la surface avec un bon degraissant...

 

Last edited: Jan 4, 2017

I have in temperature 54 c ° full load with my i7 4790k push to 4.8 mkz

http://www.tonymacx86.com/threads/mod-g5-mk-i-mk-ii-hanakintosh.193768/#post-1258282

 

I remember reading thermal paste can provide as little as 10% difference. Who knows?

Most likely it's bad cooling. What's the processor fan etc? It looks pretty packed in that case.

 

This are my fan processor :

 

I translated your post with google, maybe it was wrong translation but I understood that you are putting the paste from edge to edge... but you are not, so should be cool, don't think that is the culprit.

And whatever CPU fan are you using, it's the same fan you used on the previous CPU which was also XEON, right? So the difference, if everything else is the same, is really troubling. Except if that's a known fact that that cpu model is running that much hotter. I presume you did your research... I would return it for sure.

 

This is what I used :

1°) Thermal Material remover.

2°) Thermal Surface Purifier.

3°) Artcic Silver 5.

 

Yes you're right ,a straight line 1 millimetre wide from edge to edge as the picture here QP3:

Right the same

 

What's the old processor?

The xeon 5680 has a TDP of 130W. What's the old chip and what's its TDP?

At a guess it's 95W? If so, you've increased watts by 30% without increasing cooling. And are the fans clean on the cooler?