I've always been a big fan of reseating memory, CPU, PCI cards on systems that are shipped/moved. If temperature has remained constant but failures happen between a range of 10mins to hours suggests to me that it might not be heat related, but more probably bad memory (or badly seated memory -- DIMMs have always had a strange seating arrangement in my eyes, the connections can shake loose really easy).
But I agree, the heatsink should be just as hot as the chip, if it isn't then there is probably an interface problem -- The problem with typical heatsinks is that the Pad is meant to be "broken in" and if the heatsink wasn't mounted correctly to begin with, the pad could have "broken in" wrong, in which case it won't work well and you get the heat problems. _______________________________________________ Prime mailing list [email protected] http://hogranch.com/mailman/listinfo/prime
