• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

T-junction?

sjt269

sjt269

sjt269

Associate
Joined
11 Dec 2004
Posts
1,979
I've searched this already, but all I get are motors threads. :p


Anyways, I've noticed that Core Temp shows that I have a 'T-junction' that's either running at 85 degrees or is set to 85 degrees - if its a setting (which I suspect, as it doesn't change) then what is it, and if its a component on the motherboard, should I be worrying?

(I'm slightly paranoid about my Northbridge too, since the Freezer 7 is literally a millimetre above it, and I'm hoping it won't have any adverse affects).


As an aside, I've got no idea of my load temps (yet) but idle temps are 20 and 22 degrees on each core respectively of an E2180 @ stock. I'm leaving it a while before I start clocking it - I want to make sure the system really is properly stable. Besides, I haven't been disappointed with its performance at stock so far. :D


:edit:


Ah, and another question while I'm here. Core Temp reports my CPU frequency as '2004.71Mhz' all the time, and at '204.71 x 10' most of the time...but earlier one after un-RARing some files, it displayed the same frequency at '33- x 6'*. That looks like the computer actually overclocked itself? I dunno if Core Temp was malfunctioning (0.95.4 beta version, I think) or if the PC can actually do that. I'm aware that Speedstep and similar things can drop the multiplier to reduce CPU speeds/temps when idle, but how the hell did it increase the FSB? And more importantly, what if my RAM hadn't been capable of that setting? The PC would have randomly crashed, and I'd have been none the wiser.

*the - is because I can't remember the exact freq, but it began with two 3s. A little arithmetic would get you the exact frequency if you were all that bothered. :)



Final question - the last motherboard I had came with a monitoring software that kept a record of system temps etc, so I could look back after playing a game and see how hot it had been getting. Is there any free software that does something similar for Gigabyte boards?
 
Last edited:
Thanks for that, interesting read. No need to worry about the T-Junction, then. :p


:edit:

Actually, if I can bug you further, there is a bit that I didn't quite understand. You say that the DTS is the 'on-die' temp sensor, yet you then say that core temp = T-Junction - DTS. Why that, and not core temp = DTS, if the DTS is the on-die sensor?
 
Last edited:
Because the DTS readout is the difference between the actual CPU temp and Tjunction. Ie. the DTS will tell you how many "degrees to Tjunction".

So to find the actual temperature, you must do Core temp = T-Junction - DTS.

Also, Tjunction should not be called Tjunction, as that implies that reading denotes the temperature at the junction. It should really be called Tjunctionmax ;)

Jon

Edit: CoreTemp can log temps ;)
 
GeForce said:
Because the DTS readout is the difference between the actual CPU temp and Tjunction. Ie. the DTS will tell you how many "degrees to Tjunction".
Click to expand...

Close enough, to be strictly accurate the DTS has a negative -ve direction tending towards zero. Where the DTS value is the delta between the die temperature and the Tcc (thermal throttling point). They use Tjunctionmax in the coretemp calculation as an approximation of the unknown factory set Tcc point.
 
Thanks guys, that's two out of three queries answered. :p


I've yet to see my load temps go beyond 36 degrees in Core Temp, though I've also never seen CPU usage go above 50% on both cores, so I don't know how brilliant that is. I'm going to leave Prime on for a torture test later to see if I can get 100% CPU usage. I had a quick play with it yesterday, but like I say, it never went above 50%.


:edit:

Argh, spoke too soon. Today has seen the occasional 43, more often hovering around 40-ish than 36. :o
 
Last edited:
Ah right, I'll give that a go. I can't see version 25.2 on a quick Google, so I'll stick with whichever version I'm using now (24.14, it seems), and just run two instances. :)

:edit:

Or at least I would if I could run two instances. Clicking the icon just takes me back into the currently open one. Hmm. 25.2, here I come.
 
Last edited:
Aod said:
its not important though. they're likely to be about 10c higher, but it doesn't actually matter
Click to expand...


Which then puts my load temps up by 10 degrees and lessens the overclocking headroom. :(

Cob said:
Prime95 v25.4 :)
Click to expand...

Thanks - downloaded and will give it a shot. :)



waso_dude said:
Temps on idle sound too low tbh.

Get Everest, and speedfan and compare.
Click to expand...


Will do, but why should one software be any different than another? Surely they all read from the motherboards temperature sensors, so they should all give the same reading...right? :o


:edit:

Had a rapid change of mind on Everest when the site I was on confronted me with a purchase price. :o
 
Last edited:
Temps are reported the same on Speedfan and Core Temp, however they are indeed higher than I mentioned in the OP - partly that's because it's warmer, partly because the PC has been used fairly constantly all day, and partly because I think the temps in the OP may have been taken within 10 minutes or so of turning the PC on, so they may not be an accurate reflection of the idle temps.

Currently fluctuating around 30 degrees, depending on what I do. That sound about right? Still pretty pleased given that the rig is near enough silent, and that's without the front and rear case fans on. :p



And can someone elighten me as to what 'Vcore2' is on Speedfan? Because it's 0.7V higher than Vcore 1, and I'm hoping it's not the second core of the processor, because 1.89V sounds a little high. :eek:

Oh, and my 12V rail at 0.45V? Methinks this program is reading voltages wrong. My -12V rail is at -16.6V, and the -5V at -8.78V. Hmm. Heh, and apparently 'Temp 3' is at -2 degrees. Air cooling is pretty good these days. ;)


:edit:

That version of Prime, after a short run, puts me at 50 degrees full load, sometimes straying to 51. I imagine that would go a fair bit higher if left for longer, but I've not really got the time or patience right now. I'll maybe try it tomorrow.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom