q6600 go AND b3

[bilbonvidia]

Do the above processors have different TJ maxes? What are they and what temp monitoring program is best for each? Realtemp reports 5 C lower temps than core temp.

TJ Max realtemp =95
TJ Max Coretemp= 100

Which is more likely to be correct? Question

Also when stress testing in prime95 which is the best setting to USe, large or small FFTs?

 

[Zorg]

Yes they do have a different Tjmax. Which one do you have, G0? the Tjmax of 95C for the G0 in Real Temp is more likely to be correct, because Unclewebb did testing with an IR thermometer.

 

[bilbonvidia]

I have a GO go go! So I will assume the lower temps reported by real temp to be correct?

It's OC'd to 3.0GHz with a arctic freezer pro 7, coolermaster 360 elite case, intake and exhaust 120mm fans. Ambient around 17C core temp says around 36c idle
61 load with realtemp reporting 5c lower.

 

[Zorg]

Your ambient (room) temp is 62.6F? That's an icebox. Yes the Real Temp numbers are correct, IMO of course.

What are you using to load the CPU? If you're not using Prime95 small FFTs, then try it.

 

[bilbonvidia]

Okay that was a guess at ambient, I'll get a thermometer and get the real ambient temp.

 

[Zorg]

Not a big deal, your temps are good.

 

[sportsfanboy]

Man I'm jealous of all these low ambient temps, I hate the heat.

 

[Zorg]

I hope you were kidding. 62.6 was a guess and as cold as a witch's t... uh... never mind.

 

[pogsnet]

Maybe he is in polar region thats why the ambient is low. 17c is cool, 15-10c is for refrigerator.

 

[Zorg]

Yeah I know, but thanks for clearing that up for me.

Anyone that is in a room with an ambient temp of 62.6F is wearing a jacket and maybe some light driving gloves. :lol:

 

[bilbonvidia]

Hi again, okay I must get a thermometer!! The air temp out side was 14C at the time its 16C at the minute and I am running at 3.2 and have beeb running prime for just over an hour and temps are 63 according to coretemp and 58 to realtemp. I have been reading the realtemp thread on the other forum by unclewebb and it would seem that the b3 is TJ max 100 and the GO is 95 which would mean 58 is the correct number . Although its not definite, but aren't intel supposed to be releasing some more info todsay on TJ etc?

 

[Zorg]

You don't need a thermometer, your temps are good. So there is no confusion, ambient temp is the temp in the room where the computer is, not outside.

Do you have a link to the post about the B3 Tjmax, I was looking for it recently and couldn't find it.

Intel is supposedly releasing the Tjmax for the 45nm chips, but I don't know if they are including the 65nm.

 

[bilbonvidia]

Another stupid Q but are the cores temps usually hotter than the cpu temps?

 

[Zorg]

Tjunction (core temp) is always less than Tcase. A very good source of information is the Core 2 Quad and Duo Temperature Guide. Give it a read and it will answer your questions about the relationship between the two.

 

[bilbonvidia]

Thanks for that pointer, thats a great guide. Realtemp has changed its ideas on the Q6600 TJ max and the new beta version reports it 100 now.
http://www.fileden.com/files/2008/3/3/1794507/RealTempBeta.zip so therefore reports higher core temps but I guess the all important reading is distance to tj max anyway.

I am at 3.2 now at load with tjmax being 100 core temps reported at 62C cpu 64C. I ran prime small ffts for 6 hours without issue would you say thats good enough to say it was stable?

 

[sportsfanboy]

What does this new info have for an inpact on the B3 q6600 anyone know?

 

[Zorg]

Thanks for that pointer, thats a great guide. Realtemp has changed its ideas on the Q6600 TJ max and the new beta version reports it 100 now.
http://www.fileden.com/files/2008/3/3/1794507/RealTempBeta.zip so therefore reports higher core temps but I guess the all important reading is distance to tj max anyway.

I am at 3.2 now at load with tjmax being 100 core temps reported at 62C cpu 64C. I ran prime small ffts for 6 hours without issue would you say thats good enough to say it was stable?

Interesting, I wish they could make up their minds.

I would say your fine.

 

[sportsfanboy]

I wish they could make up their minds.

Yeah man... I finally got off my azz and calibrated realtemp so I could actually trust what I'm reading on my B3 q6600, just to be back to square 1 again, very annoying.

I'm going to say the hell will everyone, set a TJ of 120 and get phase change cooling, that will show em.

 

[Zorg]

Intel was supposed to be releasing the Tjmax at the IDF for the 45nm cores, I haven't seen any numbers yet. I hope that they include the 65nm cores as well, but I'm not holding my breath.

 

[sportsfanboy]

I'm sure when and if they do, good old Uncle web will be on top of it.

 

[CompuTronix]

Guys,

Intel disclosed Tjunction Max specifications for 45 nanometer processors only: http://intel.wingateweb.com/US08/published/sessions/TMTS001/SF08_TMTS001_100r.pdf

Desktop TJ For Dual and Quad Core CPU's

45nm Desktop Dual-Core Processors
Core 2 Duo processor E8000 and E7000 series 100°C

45 nm Desktop Quad-Core Processors
Core 2 Quad Q9000 and Q8000 series 100°C
Core 2 Extreme QX9650 95°C
Core 2 Extreme QX9770 85°C

Intel chose to ingnore the 65 nanometer community, however, the Q6600 B3, which is a pair of E6600 B2's, is Tjunction Max 85c.

Intel admits Tjunction Max has factory variations that deviate from part to part and core to core, which is why we see different core temperatures on identical hardware platforms at the same ambient temperature, and different core temperatures on the same processor. This means that if the Tjunction Max specification is 100c, some cores could be 97c while others could be 103c.

The calibration methods I developed for my Core 2 Quad and Duo Temperature Guide do not use Tjunction Max values to achieve accuracy, yet when I calibrate the Q6600 G0 on my personal rig, the results show that it has an average Tjunction Max value of 98.5c.

The bottom line is that Tjunction Max specifications are not 100% accurate, and as such, should be regarded as an approximation which always has a + / - calibration factor.

Comp

 

[sportsfanboy]

Thanks for the clarification Computronix. I thought the new info might have changed the B3 q6000 tjuntion, since Uncle web is going to 100c on some processors that were previously set to 95c.