When asked how to monitor temperatures, intel said bios

JonJ678

JonJ678

JonJ678

Soldato
Joined
22 Dec 2008
Posts
10,369
Location
England
I was hoping to get a value for tjmax for the i7 920 processor, instead I was advised to disregard software and work to the value shown in the bios. This rather conflicts with conventional wisdom, so I thought I'd post a transcript

258y16r.jpg


Please discuss

edit: If someone else feels inclined to ask a similar question, the link is here and I'd like to know what response you get.
Thread here suggests a phone call response was in direct contradiction to this. i believe the relevant number to be 0870 607 2439 however as I don't have a landline I'm reluctant to call
 
Last edited:
One thing I will say is that i have always felt that the software programs like coretemp and realtemp report temperatures way too high, on gigabytes easytune software I get far less temp reading. Another thing, I tested my old c0 chip to see how much heat and voltage it could take, i used 1.5v and realtemp said 100c instantly and thats as far as realtemp would go but it stayed stable like that for three hours with prime 95. thats why i dont believe in the software programs.

If he is from Intel he must have some credibility ........
 
I can't tell if your insulting me, bulldog or diego davetrace1, but I see no reason to consider any of these options an idiot. Explain yourself or apologise.

Not sure that's amazing advice bulldog, processors will take damage below the shutdown temperature. Still, it does suggest that worrying about temperature on these chips is probably a waste of time. That would only leave voltage and current draw to worry about :s
 
JonJ678 said:
I can't tell if your insulting me, bulldog or diego davetrace1, but I see no reason to consider any of these options an idiot. Explain yourself or apologise.

Not sure that's amazing advice bulldog, processors will take damage below the shutdown temperature. Still, it does suggest that worrying about temperature on these chips is probably a waste of time. That would only leave voltage and current draw to worry about :s
Click to expand...

I am not insulting you in any of my word. I just take intel advice. But I probably shouldn't say try 100C on all cores as it could damaged your cpu and got myself to blame. I had now removed my posts.

I better shut up and kept out of it. The choice is all down to yourself to kept realtemp as a safer way.
 
Ah, what a mess. The challenge was aimed at davetrace1, not at you bulldog. He's calling someone an idiot and I can't work out who. Plus your first post (sadly deleted :( ) was rather good, it was the second suggesting everyone load their processors to 100 degrees to see what happens which was a bit dubious :)

I'm tempted to take Intels word on it. If it fails thermally while never displaying over 67.9 degrees in the bios, the rma should go smoothly with a copy of the transcript. If it lasts me for three years running peacefully at very high temperatures I'll be thrilled.

I'm sorry this thread has been so incoherent
 
Upon re-reading that's my best guess as well. Calling intel's technical support stupid makes absolutely no sense to me though.

Here's hoping this thread becomes something more than a long list of -post delete- with a locked button at the bottom, I hadn't seen that intels thermal spec applies to the bios and not to software readings before so it seemed a useful thing to post. If it's common knowledge I'll be suitably embarrassed
 
It's known that the maximum i7 temperature is supposedly 67.9C (like any BIOS temperature reading is that accurate anyway) but we also know that it throttles at ~95C core temperature in order to not go above 100C core temperature. So it's still sensible to use the core temperature as a measure for whether an overclock is safe or not.

Just because they don't want to officially disclose the tjmax doesn't mean everyone doesn't know it (or have a good guess based on other info).
 
That's a good point. If tjmax is the temperature it shuts down at, then turning off the fan/pump, loading prime and watching hw monitor until it halts would give a reading for tjmax anyway

Is the conclusion then that this entire thread is a waste of time? I fear it may be so
 
JonJ678 said:
Is the conclusion then that this entire thread is a waste of time? I fear it may be so
Click to expand...

I dont conclude that, if anything its an interesting read. I have always wondered if these software temperature programs are accurate and the fact that someone from Intel has said this is significint because they cant just say anything. I had a realtemp saying 100c on my old i7 c0 chip when i did a suicide run and my system never shut off even though realtemp said it was a tjmax.
I hope more people will post with more info.

My processor temprature reads 32c in the bios, so like you say JonJ678 if the intel guy is right, its voltage we need to worry about more then heat.

32 does seem unbelievably low though,
 
Where in the BIOS could you see the individual core temps? I haven't seen a BIOS yet that gives core temps.

You would think that the BIOS would display the most realistic temperatures, no OS or software getting in the way. The BIOS is the hardware/software interface and therefore should be more accurate.

We've all used realtemp, coretemp, everest etc, but that is software, and that has to communicate with the hardware - through the BIOS.

In my experience with temp software it always seems to show at least 5C less than the BIOS. On a few of my AMD systems it could be more like 10C. Then there is the temp monitoring software that comes with you're motherboard, not good for anyone suffering from a nervous condition.

To get totally accurate temps you need some serious and expensive equipment. If you use software for checking temps, which we all do, best just to use it as basic guide. Its the best we've got, unless you have the keys to a lab at MIT.

Back to Diego and temp spec of 67.9C. Where did the .9 come from? What did he take this measurement with? What are the tolerances involved? Is this the same spec for all i7s? The list goes on...

Sounds like Diego is reading 'the customer service guide to the i7' directly from the screen in front of him. It's when they say things like 'you're call is important to us' you know you're talking to a robot reading from a list.

Wouldn't it be nice if Intel had an overclocking forum. That would be interesting. Do you think they would go for it?
 
Last edited:
Why does everyone suddenly seem to agree that running a 45nm i7 CPU at 85c is fine?

When running ANY other 45nm cpu at over 70c is a bad thing and MUST be avoided?

Not sure where this temp tolerance came from...
 
no because as soon as you overclock Intel state that you are no longer under warrenty!

Intel do give out some strange advise had some intresting email contact with them a while back, will see if i can dig it out.
 
Sayso said:
no because as soon as you overclock Intel state that you are no longer under warrenty!

Intel do give out some strange advise had some intresting email contact with them a while back, will see if i can dig it out.
Click to expand...

This is right and correct, once you overclock your cpu processor on intel, your warranty will be no longer in use, (only the normal default intel specs will carry the full warranty) but lucky for us all, cos our OCUK i7 920 4Ghz bundle still under 1 year warranty with OCUK not Intel.
 
You must log in or register to reply here.
Back
Top Bottom