Possible Pandora Polling Preventative Problem Protection


atomicthumbs

I am the king of the collectors, with hospital bed
Joined
Apr 17, 2006
Messages
2,827
Age
33
Location
XK-Masada
Website
printeronfire.org
Can the temperature sensor of the OMAP3530 be read in hardware?

If so, I suggest that the Pandora OS be changed so that overclocking the processor makes it start polling the temperature sensor. If it went outside the operating specs (~90C, or less to be safe), it could shut it down or downclock it.

The implementation of such a thing would involve such words as "kernel" and "driver".

If we did this, and if its possible, I can see it preventing a lot of problems (if overclocking turns out to be problemative).
 
90C? If it ever gets that hot, I don't want to be holding it! My laptop peaks at less than that and still manages to burn me.

I don't think overheating could possibly happen before the CPU just dies of chronic confusion, but there are already threads open trying to ascertain whether or not this is even going to be a problem. Probably best to work out if it *is* a problem before trying to solve it.
 
WoD said:
90C? If it ever gets that hot, I don't want to be holding it! My laptop peaks at less than that and still manages to burn me.

I don't think overheating could possibly happen before the CPU just dies of chronic confusion, but there are already threads open trying to ascertain whether or not this is even going to be a problem. Probably best to work out if it *is* a problem before trying to solve it.
OK, you cannot compare a x86 device against ARM (i dont know how many many times it's been said).

also we have to talk about processor core temp. VS, external case temp.
 
Last edited by a moderator:
DroneB Dev said:
WoD said:
90C? If it ever gets that hot, I don't want to be holding it! My laptop peaks at less than that and still manages to burn me.

I don't think overheating could possibly happen before the CPU just dies of chronic confusion, but there are already threads open trying to ascertain whether or not this is even going to be a problem. Probably best to work out if it *is* a problem before trying to solve it.
OK, you cannot compare a x86 device against ARM (i dont know how many many times it's been said).

also we have to talk about processor core temp. VS, external case temp.
I think when discussing temperature it is totally fair to compare x86 against ARM: 90C is 90C no matter what device it comes out of :P
And given how thin the case is and how packed the entire thing is, I would wager that the CPU temperature would be nearly the same as the case temperature underneath it anyway. There's no heat sink to direct the heat outward like there is in an x86, so it just dissipates through everything it contacts.
 
Last edited by a moderator:
Back
Top