Originally posted by: Cherub
I noticed that after upgrading from Everest 4.00 to 4.20, it now reads the cores separately from the "CPU" and each is about 10c higher. Also, Everest began reading my GPU about 8-9c lower than it had with 4.00, and lower than AtiTool reads it. CPU-z also does not read the vCore; it always reads 1.150v no matter what the actual vCore. I think the software is more problematic than the diodes, due to variations in what is being reported by different software, but that's just a guess.
Well, some (as I said) 8400's work
as expected with all the various utilities, old and new..
So your theory goes right into the toilet
Based on what you said, yours is working properly, and the later everest is now giving updated info for the chip itself.
Peeps with far more knowledge than you or me have noted the prob - if you actually research some of the google linked articles.
While there might be some detection variation due to the chips newness - since the software writers lag a bit on new CPU versions - there is
no explanation for a chip temp being stuck at 37 degrees permanent, or 14 degrees dif in temp of each core idling.
FWIW, coretemp is the most accurate reader of "real life" temps, per many threads on XS.
http://www.alcpu.com/CoreTemp/
Whats missing from this picture is folks noting WHICH batch codes are "good" and which are "bad". Or perhaps there are good and bad within batches.
It would be informative if the buyers here reported on their outcomes.
But if you install a known good Core2Duo HSF, there is really nothing to worry about, really - heh