Ambient On Air!!

[Zardokk]

Hey guys, take a look at this:

http://www.kingsasquatch.com/ambient_on_air.jpg

That's right, 21C on core 1. It actually fell down to 20C at one point. What's amazing is that the ambient temperature in my basement is 20C! Plus, this is just under a Thermalright SI-97A! Now yes this was an idle temp, and I know idle means nothing, but come on...gimme some credit. I hit ambient temp ON AIR by doing nothing special...IHS not removed/lapped/anything. So...how'd I do?

[saber63]

your cpu temp moniter and what you used to moniter you ambiet arnt calbrated to the same degrea,(ie one or Both are either a degrea to warm or cool) sorry its more of a bad sign if anything

levi

[Zardokk]

I don't think it's a bad sign. Yes I'm probably 1-2C away from true ambient, but that's not the point. This is one of the lowest readings I've ever seen on air, and it's damn near accurate. Core Temp is very, very close to actual temps, closer than any other temp monitor I've come across. And the thermometer in the room is reading at 68F, which is almost exactly 20C.

[Revv23]

si-97, isnt that a p4 heatsink?

[Paapaa]

Lol, so the core2 is at 30?? Maybe you should wait until the BETA status goes away...

[Zardokk]

That's actually acurate. Once core is significantly hotter than the other. It's probably due to the fact that my heatsink doesn't completely cover the IHS. But Core Temp is very accurate and core 1 really was near ambient. Woot woot!

[Serra]

Core Temp is very, very close to actual temps, closer than any other temp monitor I've come across.

Even though you're clearly idle, I have a very strong feeling you not at ambient. I base that on something called "logic". There have been accounts of cases where Core Temp reads thermal data just plain wrong, and I think this is one of them.

Why?

If your core to truly be at 20 degrees, let's think of what has to go on. Well, logically the infallability of Core Temp (or so people seem to think - Core Temp is not the end-all be-all of temperature monitoring) would state that the temperature in the center of your processor is 20, and since that is where heat is produced, there must be something of even lower heat to draw that excess energy away very quickly to not get more less than 1 degree difference. What does that mean? Well, it means for one thing that for you to ever reach ambient your heat sink itself HAS to be at BELOW-AMBIENT temperatures. But wait, there's more! Given the fact that the temperature inside your case is SURELY a degree or two higher than that of your room, the 'ambient' temperature for your case would be that degree or two (or more, I've opened a lot of cases where I'm sure the difference was 10+ degrees)... and your heatsink could not possibly reduce it below that level. Another impossibility in the world of physics. You could help negate this by having an open case or a LOT of airflow, but if the temperature inside is even 1 degree off... well, again, impossible.


So in answer to your question: you didn't!

Edit: :soap:
Core Temp is not the end-all be-all of temperature monitoring! No offense to member "The Coolest" whom I'll be quoting in this rant, but despite his claims that "The temperature readings are very accurate as the data is collected from a Digital Thermal Sensor (or DTS) which is located in each individual processing core, near the hottest part.", his statement is just wrong. Core Temp is perhaps marginally better than Everest or SmartGuardian (not sure, but I'll allow it might be), but it would only be so by making guesses about many things. The *ONLY* way to make an actually accurate temperature monitoring program would be to take an actual thermal probe and wedge iit into various CPUs on various motherboards and collect data about the delta between what is software displayed and what is actually there for each CPU and/or Motherboard... which is not going to happen. I did once see a project that looked like it had some promise where people would take temperature data and incorporate voltage, current draw, and efficiency and tried to base a program on that... but it turned out the guy in charge was pretty well just pulling efficiency numbers out of his hat for almost every CPU and so that project collapsed.

I have absolutely no problem with people using a software monitor to help them find DELTAs between their setup with heatsink Y and heatsink X, but to claim the results are real-world accurate is just ridiculous.

Edit II: Actually, just check out the thread about Core Temp by its creator (or who I think is the creator) on XS:http://www.xtremesystems.org/forums/showthread.php?t=103638 you'll note that there are is a large number of awful, clearly wrong temperature data posted in that thread (and yet people still rant about how awesome-azing it is).

[Zardokk]

I'm not saying that Core Temp is something great and fantastic. I'm just saying that I got closer to ambient then ever before according to core temp. I was probably more realistically 2-4C off of ambient, but I was still closer than ever. After thinking, I realize that it was in fact pretty illogical to think that I had reached ambient temps, but heh...I got closer than usual.

EDIT: The temps read from Everest match up perfectly with the temps in Core Temp. The temp reporting is accurate within 1 or 2C. My ambient temp could've been lower than my thermometer was reading, though, which could've easily been one of the issues here. Core #1 typically sits at 25-26C and core #2 is usually about 32-34C. Could've been a change in ambient or something...dunno...it is cold outside and people do go in/out of the house a lot.

[ragnarok95]

si-97, isnt that a p4 heatsink?

Its SI97A.

For 939 socket.

[HaxR3]

Meh, ive seen 14*C CPU temp on air

(my whole pc had been in a cold car boot for the whole day)