Yes, but your at a 15c ambient, fully loaded 30-40c-ish with a 15c ambient sounds about right. Your not stressing the CPU enough to push it over where the sensors stuck I'm sorry to say. If the diode is STUCK at 33c, you need to load it(heat it up) PAST where its stuck. With a room ambient of 15c, your NOT doing that, even when fully loaded.Yes, always at 33 for both cores at under load AND idle. :blarg: I'll try 4ghz in a bit.
Read his post again, his ambient is 15c. Hes got a STUCK diode, so it doesn't read below 33c period. You COULD have LN2 at -112c, but his CPU would still read 33c regardless because its stuck. Only way to get it "unstuck" is to heat it up PAST the "stuck" temp. The delta is unknown.There is no way there would be no temp delta from idle to load, LN2 and the best designed LN2 pot ever made cant even make delta 0. Ive never heard of a chip not giving temp readings, I would bet it is the motherboard. What board is it?
Thats the thing, the diodes are only stuck BELOW the "stuck" temps, once the temperature rise's 5c or so above the "stuck" level, they work normally. Thats why I want him to test is in normal ambient(20-25c) environment. The diodes are known to do this(get stuck) in the 45nm CPU's, but it doesn't prevent them from working normally/doing their job, which is underclocking/shutdown when getting close to TjMax.That should be an issue for RMA. The thermal shutdown protection works off those diodes, right? What happens if his cooling solution suddenly failed? Granted, the stock heatsink MIGHT be able to keep things under TjMax, even without the fan spinning, but not necessarily.