-
This is all nice theoretical stuff but we still don't have a clue then how hot the cores are, the developpers need to get proper input from intel to get their software reading out the cores correctly or at least be able to give us an idea if we are still in the safe zone... but thing that worries me the most is that some get let us say real values while others are really messed up: too much difference between the cores, idling at sub ambient temps,.... so there might be more here going on then we can imagine...
Time to call Mulder and Scully !! and maybe Mc Guyver too
-
anyone tried a temp probe or gun and compared to their older E6 series?
-
I've got the hardware to both use a temp-probe or gun to measure temps but don't know how to properly take the measurements? It sounds easy but how do you go about using say a "K-Type" probe or non-contact probe to measure the CPU temps? I thought about maybe drilling a small hole perpendicular to the bottom of my CPU waterblock and mounting my temp-probe into the hole using some TIM to fill any voids? Although that might work somewhat it certainly won't be accurate because of the material between probe and CPU, plus I don't know how deep I'd be able to drill into my CPU block (making some type of fixture to hold block so I could drill using mill or drillpress would be difficult at best? If anyone has any decent ideas please share them!
-
1 Attachment(s)
1)underclock and undervolt cpu and be quick or hold your block against and remove and measure, but if undervolt enough, ok with nothing (when later increasing back to normal, increase volts first, save, then increase mhz...or enjoy reboot hell.)
2)take block off and put piece of masking tape over back of cpu (temp is inaccurate with IR gun off shiny back of cpu without tape...learned that from unclewebb post)
3)start computer
4)I positioned screen so that cpu temps and IR gun both visible in pic....temp 2 is cpu temp.
this is my E6850. The cpu (tcase) temps were very accurate in software (temp 2), comparing reading to laserpro temp gun (calibrated accurately with very accurate temp probe).
Note the core temps are off 15C, as this test was done before it was known that tjmax was 100c, so should read 75 and 74C.
I have done the same with my E8400, cpu temp reports same as IR gun at 50C and 60C. Need two people for pics, so no pics of E8400. Who knows about core temps.
btw...when i had looked this last year...tcase sensors that listed as cpu temp are supposedly fairly accurate (always exceptions/bad sensors), within few C, especially between 50 and 80C, where they are calibrated for...listed in some chart in intel specs...too long ago to remember...and to lazy search again for.
-
http://www.xtremesystems.org/forums/...d.php?t=179044
real temp program WORKS FLAWLESSLY stock e8400 ubder water now reads right 34-30
