I'm running a duron 700 @ 1000mhz, and I'm curious about something, I asus probe 2.12 was reporting my cpu temps around 48-50c degrees idle, and it was dogging my machine down for some reason. I removed it, and installed 2.11. now heres the wierd thing, 2.11 reports my cpu to be around 38-40c idle, and around 45-47c under full load for 6 hours. What I want to know is, which version is misfiring it's reports?
Heh! There's a post somewhere in here within the last couple of days. Seems that two different people had the exact opposite experience with the different versions, i.e., one found 2.11 was inaccurate and the other, visa versa. You could always try comparing them with the BIOS temp (however that is measured) or install Motherboard Monitor (if it supports your board) to see if it's closer to one of the Probe versions.
the 211 is closer to bios, but the temps weren't the reason I changed, 2.12 was causing my system to slow down to 386 speeds, even when it wasn't overclocked (when I first installed 2.12, I had already bumped it up to 1000mhz, and I thought at first that it was a bad overclock). 2.11 however doesn't cause me any problems. thanks for the info.
my first a7v board (the one I was writing about) was v1.01, and it had the little temp lead sitting on the cpu up against to the core, Yesterday, I got another a7v rev1.02 and it has the thingy (strong technical term, some may not understand eheh) in the middle of the socket. the 1.02 board shows the temps to be about 10c cooler than 1.01, same bios rev, which pretty much confirms what I've been reading about the kt7 boards, the temp isn't accurate because the lead isn't sitting on the core. I hooked up an external monitor today, and found to my suprise, that the probe 2.11 is about as accurate as you can get without having a temp gauge inside the core itself. anyway, the cpu temp is around 46-50c at 1000mhz (duron700). a7v rev 1.02 reports between 35-44 at the same speed.
January 13, 2001 2:08:24 AM
btw both of the temps at the end of my last post were at full load. thank you, drive through.
January 13, 2001 3:43:52 PM
To assume that these temps are as good as internal diode temp is a bit off. Even an probe touching up to the core isn't going to be quite as high as true CPU core temp, due to the fact you're measuring a secondary heat pathway(albeit not as horrible a secondary heat pathway as the backside).
Low kt7 temps are normal even when the thermsitor(or socket-thermistor) is touching the CPU backside. A lot of people mistake this as cpu core, but it isn't. It is a ceramic substrate, wtih high thermal resistance. It is due to this thermal resistance that any temps measured on the backside of a FC cpu do not represent full core temp changes, and are highly inaccurate(and often too low).
as I said (the part I guess you failed to read) about as accurate as you can get >>without<<--!!keyword here!! having a temp gauge inside the core itself.
not going off on you or anything, I just didn't say what you are implying. I realize that I'm gauging a secondary heat path, I'm just saying that the asus probe 2.11 measures temps as well as an external sensor, and that 2.12 seems to be way off. thank you.