coretemp 70C = CPU too hot?

[Jerry McAllister]

On Fri, May 09, 2008 at 08:10:40AM +0200, Christian Zachariasen wrote:

> On Fri, May 9, 2008 at 4:55 AM, Nerius Landys <nlandys at gmail.com> wrote:
> 
> > Howdy.  I purchased a 1U 10 inch deep server machine a few months ago:
> >
> > http://www.abmx.com/1u-10inch-deep-supermicro-mini-server-p-366.html?osCsid=80f3951929d5a7ae27a51733627ee18a
> >
> > The CPU is a Xeon 3xxx dual core 2.4 GHz.  The machine has no case fans, by
> > design.
> >
> > It's sitting in a well-ventilated rack in a data center.  Oddly, when there
> > are no machines below and above it, the machine gets hotter.  Seems that
> > machines above and below help to cool it down.  I have the "coretemp"
> > kernel
> > module loaded on the FreeBSD 7.0 OS, and I saw that the CPU core temp(s)
> > hit
> > 70 degrees Celsius during a compile of GCC.  Is this too hot?  Should I
> > complain to the people who assembled the computer?  At the time this
> > happened there were supposedly no surrounding machines.  This machine has
> > given me no problems.  At idle when conditions are good (meaning A/C is
> > working properly and there are machines above and below it) my CPU temps
> > are
> > below 40.
> > _______________________________________________
> > freebsd-questions at freebsd.org mailing list
> > http://lists.freebsd.org/mailman/listinfo/freebsd-questions
> > To unsubscribe, send any mail to "
> > freebsd-questions-unsubscribe at freebsd.org"
> >
> 
> As with so many other things in the computer world, it depends. With no case
> fans, it's weird that the computer gets colder if it has something above and
> underneath it. But there are so many factors when it comes to case
> temperatures and air flow that it's nearly impossible to tell why. If
> there's a huge cooler on top of the Xeon (Processor wind tunnel), then it
> could be that closing the ventilation holes in the top and bottom of the
> case makes the air flow more directly from the front to the back of the
> case.

Could be or some other bernouli/venturi effect.
Many systems will get hot when the cabinet is opened because
it breaks the airflow path in some way.

////jerry


> 
> It seems the Xeon shuts down the system automatically if it reaches 105 C. I
> don't know if this is any pointer to what a reasonable 100% load-temperature
> could be, but I know processors nowadays run much cooler than they used to.
> (I'm used to AMD Athlons on or above 70 C idle)
> 
> I'd say you should be fine if you haven't seen any instability at 70 C with
> 100% load.
> 
> You could try this if you want to be sure:
> 
> ---
> On Tue, Feb 28, 2006 at 12:27:07PM -0800, Don O'Neil wrote:
> > What is the best way to 'burn in' or 'stress test' a new system w/
> FreeBSD?
> > I'd like to stress test the CPU, Memory, Disk, etc.. To make sure the
> > hardware is 100% good before putting it in production.
> 
> Doing something like a buildworld -j64 loop (if you have enough
> memory, otherwise reduce -j level to avoid swapping) is going to
> exercise your system a fair bit.
> 
> Kris
> ---
> 
> 
> Regards,
> Christian Zachariasen
> _______________________________________________
> freebsd-questions at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-questions
> To unsubscribe, send any mail to "freebsd-questions-unsubscribe at freebsd.org"