So, are we running a system that can suddenly crash ?!
(searched on deja/google about PIV problems with 5.0.5, but found
Any hint deeply welcome.
It *will run on OpenServer, but not on 5.0.5; You need OSR 5.0.6 with
Any release of OpenServer _should_ run on a Pentium 4, but it might
be dangerous. rs506a turns on some of the thermal throttling and
safety features, and modifies how some loops are done in order to
avoid triggering those features in the first place. Without these
modifications, under extraordinary circumstances the CPU could overheat
and either shut down, or possibly even damage itself.
I believe the sort of circumstances needed are along the lines of: the
CPU fan broke, and some part of the kernel got into an infinite spin
loop (or at least did a tight spin loop for several minutes).
The level of danger would also depend on exactly which CPU is being
used, at what clock rate, with what other passive cooling mechanisms in
the box. I believe the newer P4's are actually fairly low-power and
thus dissipate a lot less heat == much less danger of overheating.
According to this summary article http://www.overclockers.com/articles517/
which references original Intel docs, there should be no more danger running
a P4 without special thermal management code than running a PIII or earlier
processor. If things get really out of hand (such as the heatsink falling
off), the processor will halt before damage occurs. With thermal monitoring
code enabled by the BIOS, the processor will attempt to throttle back the
clock to reduce thermal output. But this thermal monitoring/throttling
circuit is independent of the THERMTRIP circuit.
This is not the case with AMD processors, which spectacularly self-destruct
when the heatsink is removed as documented a few months ago at
Does this also apply to Intel's Xeon chip (5.0.6 and not 5.0.5)???