Real Temp - New temp program for Intel Core processors

[karateo]

I had a look at things and the only thing that I noticed that might be causing problems for users is if the driver is not being loaded properly then the program won't be able to read your temperatures and will display 100C. I added some code to prevent the program from starting up if it has any issues with loading the driver.

The driver might not load in Vista if you are in a limited account.

This is my best guess. Version 2.1 is now available if anyone having problems wants to try these theories.

http://www.fileden.com/files/2008/3/...7/RealTemp.zip

error in vista business 64 bit
Driver error
Driver not found

either normal or administrator user

[theonlybabyface]

error in vista business 64 bit
Driver error
Driver not found

either normal or administrator user

Doesn't work in 64 bit yet Karateo....read the entire post before posting.

[unclewebb]

x64 support is still in the future. I need to make sure that this new driver is working 100% in 32 bit before trying to make it work for x64.

[Ace-a-Rue]

you are doing a DARN GOOD JOB!...many thanks!

[Brama]

Under Vista 32, starting RealTemp 2.1 normally, I have the error "driver not loaded"

Starting it as administrator, it works perfectly.

[DOM.]

Dom7184: Have you ever tried doing the low temp test where you run at about 1600 MHz and 1.10 volts and check your idle temps. I have a B2 E6400 just like yours and I found that my computer would report below ambient idle temperatures during this test. The ++ idle temp correction adjusts my idle temperatures on that processor back up to reality.

now im getting 1-1 also with both 2.0 and 2.1

[Brama]

If your idle temps are different between your two cores by 20C then that is a big red flag that core1 is stuck. What does the Test Sensor test show?

At my idle situation (2580 MHz and 1.36 V.....I know, not a minimal idle as you suggested) I have my E8400 at -4 C° and +27 C°; at 100% of both cores (prime95) I have 22 C° and 42 C°.

Testing sensors with realtemp I have them moving 18 and 10.

It seems my sensors are not stuck and probably at TjMax they will measure the same data; the slope of the linear curve seems a lot different between the two sensors and if I could adjust the two cores separately, probably I can compensate the different slope.

Any comment?

[ixtapalapaquetl]

There is an interesting submission at Anandtech.

More than a few programs have been released over the last few years, each claiming to accurately report these DTS values in real-time. The truth is that none can be fully trusted as the Tjunction values utilized in these transformations may not always be correct. Moreover, Intel representatives have informed us that these as-of-yet unpublished Tjunction values may actually vary from model to model - sometimes even between different steppings - and that the temperature response curves may not be entirely accurate across the whole reporting range. Since all of today's monitoring programs have come to incorrectly assume that Tjunction values are a function of the processor family/stepping only, we have no choice but to call everything we thought we had come to know into question. Until Intel decides to publish these values on a per-model basis, the best these DTS readings can do for us is give a relative indication of each core's remaining thermal margin, whatever that may be.

More info here: http://anandtech.com/cpuchipsets/int...spx?i=3251&p=4

[clicker666]

Hmmmm.... I'm not entirely convinced my TjMax on my E8400 is 95C, although I'm more than willing to be told otherwise. I have a CoolIt Elite (and am soon moving to watercooling due to the noise) and made the following observations.

I tried cranking the MTEC to max cooling by dropping the desired coolant temp to 10C. My ambient is about 17C. Hitting a coolant temp of 21C was able to bring a CPU temp average across the cores of about 23C. I'm thinking that at idle that seems pretty accurate. Since I am under the assumption that the coolant temp is measured internally on the cooler itself, that would indicate the TjMax would be correct at about 105C, would it not?

The only reason I'm thinking this is because I'm set at ++ to get a temp that seems in line with the coolant temp at idle. Anything less would give me a temp below coolant temp - which shouldn't be possible. Once again, I am assuming that the coolant temp is measured within the Elite itself as opposed to any external sensors.

[jason4207]

now im getting 1-1 also with both 2.0 and 2.1

I don't think your supposed to do the sensor test at the low voltage. That is to determine your offset. Look at your idle temps and compare them to ambients. If they are too high or low use the -, --, +, or ++ setting to get core temps w/n 3-5* above ambient.

There is an interesting submission at Anantech.



More info here: http://anandtech.com/cpuchipsets/int...spx?i=3251&p=4

Just read it from a link over on OCF. Was going to post it, but you beat me to it!

That article kind of throws a monkey wrench into this whole matter! If the article is correct we are back to square 1, and the only advice I can see is to just keep your distance from Tjmax...whatever that may be.

Forget about absolute temps unless you have an IR thermometer or temp probe.

[unclewebb]

There is an interesting submission at Anantech.
More info here: http://anandtech.com/cpuchipsets/int...spx?i=3251&p=4

Is this really a sign that someone at Intel has been checking out this forum and discovered that we're way too close to the truth now so it was time for a PR to try and throw us off the trail! Maybe Intel isn't very happy with software that shows that their DTS sensors sometimes get stuck on their new 45nm chips. This wasn't an issue before with 65nm.

[unclewebb]

now im getting 1-1 also with both 2.0 and 2.1

That is normal too. When you drop the volts and MHz down, the processor does not increase in temperature very much at all, even at full load. I was hoping you could report what your reported temperature was vs your ambient temperature and how they compare. I'm hoping that at an Idle setting of 0 with low volts and low MHz that your reported temps will go below ambient, like many other B2 processors do.

[DOM.]

That is normal too. When you drop the volts and MHz down, the processor does not increase in temperature very much at all, even at full load. I was hoping you could report what your reported temperature was vs your ambient temperature and how they compare. I'm hoping that at an Idle setting of 0 with low volts and low MHz that your reported temps will go below ambient, like many other B2 processors do.

okay i check , but does it have to be 1600Ghz or can it be less ? I can get it to 600, 700 or 800 or stock on the lowest volts ?

[Qbical]

so it seems to me like my core that is cooler is prob. more accurate.
anybody?

[unclewebb]

Dom7184: If you can run at 1600 MHz then show that one just so I have something I can directly compare to. If you can go lower then post that info as well. The more data the better!

Moreover, Intel representatives have informed us that these as-of-yet unpublished Tjunction values may actually vary from model to model - sometimes even between different steppings - and that the temperature response curves may not be entirely accurate across the whole reporting range.

Isn't that the conclusion I came to in post #1? Interesting that Intel has never, ever said anything publicly about this subject until my chart showed the real truth. Remember, you read it here first at XS!

[DOM.]

room temp 75F 24C Humidity 27%

[lasershock]

lasershock: Did previous versions work fine for you?

I'm running 64-bit Vista so I haven't been able to run the previous versions.

My batch is Q746A502 btw.

[HDCHOPPER]

thks for the answer unclewebb

[ixtapalapaquetl]

Isn't that the conclusion I came to in post #1?

My thoughts exactly.

[Ace-a-Rue]

Is this really a sign that someone at Intel has been checking out this forum and discovered that we're way too close to the truth now so it was time for a PR to try and throw us off the trail! Maybe Intel isn't very happy with software that shows that their DTS sensors sometimes get stuck on their new 45nm chips. This wasn't an issue before with 65nm.

i think it is more about many buyers returning their penryn/wolfdale processors to their online retailers because of the bum DTS sensors...i did it for a QX9650...

[Ace-a-Rue]

COOLEST updated his program, coretemp, to work in vista64:

http://www.xtremesystems.org/forums/...37#post2819237

[Retrospekt]

I am definitely getting confused here...

I have an e4300 at 1.4v and 3.0ghz and CoreTemp shows me 55c load. RealTemp shows me 40c load. A whole 15c gap? What's up with that? If it's correct...I'm pushing this thing farther.

[jason4207]

I am definitely getting confused here...

I have an e4300 at 1.4v and 3.0ghz and CoreTemp shows me 55c load. RealTemp shows me 40c load. A whole 15c gap? What's up with that? If it's correct...I'm pushing this thing farther.

What are the 2 reported Tjmax temps? If one is 85, and the other is 100 then there is your answer.

[Retrospekt]

Yeah, CoreTemp says TJmax is 100, but RealTemp says 85. I assume RealTemp is wrong then? Sorry, I'm a noob.

[Ace-a-Rue]

probably realtemp is 'more' accurrate...for now, i will use coretemp, since i am using vista64, until unclewebb is able to get his program to work in vista64..