RealTemp General Discussion

[burebista]

Why not?

Because is useless. It's an arbitrary value calibrated in BIOS by every MB manufacturer.
All that matter for your CPU thermal behavior (throttle/shutdown) are core temperatures. In fact all you need to know is distance to TJMax because when your CPU reach TJMax value it will be automatically throttle itself and when it will go beyond that TJMax value (with ~30 units) it will be shutdown itself.
That's all. No need for a generic CPU temperature.

 

[ejme]

Thank you for the input. Next time please refrain from using the word "meaningless" when posting replys. It is rude, insulting and degrading implying that one is less of a person because of a simple inquiry deserving an informative reply which you think is "meaningless" contrary to the
fact that it is what people refer to as simply 'knowledge' . Again, thank you for the input.

 

[burebista]

Sorry man, English is not my native language as you can see from my location.
I didn't want to be/sound rude. Here in this part of world we are more relaxed when we speak in forums.

And maybe next time I won't post a reply, I'll let the guy figure himself.

 

[pook]

Realtemp GT doesn't seem to work well with my LGA2011 (Sandy Bridge E) chip. Is there an update planned? It only shows 4 cores.

 

[95Viper]

This a little off topic, however, I had to post this rebuttal in defense of a great poster and forum member - burebista.

Well, RealTemp displays only core temperatures (the one that really matters).
Why would you need to know a meaningless value like CPU temp?

Because is useless. It's an arbitrary value calibrated in BIOS by every MB manufacturer.
All that matter for your CPU thermal behavior (throttle/shutdown) are core temperatures. In fact all you need to know is distance to TJMax because when your CPU reach TJMax value it will be automatically throttle itself and when it will go beyond that TJMax value (with ~30 units) it will be shutdown itself.
That's all. No need for a generic CPU temperature.

Thank you for the input. Next time please refrain from using the word "meaningless" when posting replys. It is rude, insulting and degrading implying that one is less of a person because of a simple inquiry deserving an informative reply which you think is "meaningless" contrary to the
fact that it is what people refer to as simply 'knowledge' . Again, thank you for the input.

Burebista was using the adjective in a question, expecting an answer. He, also, clarified his statement and question in an additional post.
IMO, you took the statement/question the wrong way and, also, believed it was directed at you personally, by your post.
I am not berating or castigating you, ejme... I am just letting burebista know his use of the English language was correct and proper; so as, he will continue to post and help the ones in need.

meaningless value like CPU temp.

He was stating the the value was meaningless, not the person posting and made no allusion to such.
An adjective is a 'describing' word; the main syntactic role of which is to qualify a noun or noun phrase, giving more information about the object signified.

Sorry man, English is not my native language as you can see from my location.
I didn't want to be/sound rude. Here in this part of world we are more relaxed when we speak in forums.

And maybe next time I won't post a reply, I'll let the guy figure himself.

Don't let it put off, burebista; you used the adjective correctly...

Definition of MEANINGLESS

1: having no meaning; especially: lacking any significance

 

[unclewebb]

I am just in the process of updating RealTemp GT for the new Sandy Bridge E CPUs. The new Sandy Bridge E CPUs are organized a little differently internally compared to the previous 6 core CPUs. It should be ready for testing in the next day or two.

 

[pook]

Cool. Thanks.

 

[unclewebb]

RealTemp GT 3.66
http://www.mediafire.com/?ns1g4y2hinoa977

This version should work in theory for the new Sandy Bridge E series. I don't have a 3960X to play with so let me know if you find any problems.
I also fixed the log file formatting so it should be a little easier to read.

If you don't have a 6 core CPU then run RealTemp.exe instead.

 

[pook]

Yep.. that did the trick.

Thanks

 

[burebista]

Viper thanks, that's exactly what I want to say in my posts.

Kevin thanks again too for updating your tools.

 

[LocutusX]

I've got a 2 year old i7-920 here. RealTemp 3.67 is working fine. Any reason to upgrade to the latest version (appears to be 3.691)?

 

[stasio]

I've got a 2 year old i7-920 here. RealTemp 3.67 is working fine. Any reason to upgrade to the latest version (appears to be 3.691)?

Sandy Brigde.

 

[EarthDog]

Works great for me UW... thanks so much for continuing to support this product!

 

[iopq]

I'm not quite sure how to interpret the results
what's the movement test and why are the results different?

does my CPU seem like it's running too hot? the ambient is 20C, the idle temp went only to 36C
I'm using a coolermaster heatsink that's like more than 1 pound, I would have expected much better temps

I am using SpeedFan to control the fans, though so it's set between 30% and 85% (don't want it to be too loud)
what's weird is that in normal usage at 100% CPU usage core 0 is 4-5 degrees hotter than core 1, but in the test they are equal

anyway, I would like to know what temps I should run my processor at if I'm going to be running it at 100% load 24/7 (for example)
why I'm doing the tests anyway is to set good fan-temperature curves in speedfan

 

[burebista]

CPU lod 101%. I'd keep this screenshot.
Your temps looks fine but for stock speed a little bit on the high side though.

 

[iopq]

CPU lod 101%. I'd keep this screenshot.
Your temps looks fine but for stock speed a little bit on the high side though.

what temps are safe to run my CPU on 24/7? I could set it at 100% CPU use or 50% CPU use (1 thread)

 

[LocutusX]

Sandy Brigde.

OK, so I guess since I don't have a Sandy Bridge processor, no need to upgrade?

BTW just out of curiousity, would this app work with a Netbook which uses the ATOM processor? Got a netbook for dirt cheap recently, would be neat to check out temps on that too.

 

[Dauntless]

Incorrect TJMax value

Hi,

I'm using RealTemp to monitor the temperature of my Atom D425

The program works great and does everything I need it to, but it seems to be reading the TJMax value incorrectly at 90°C

The TJMax value of my Atom D425 is 100°C as you can see here:
http://ark.intel.com/products/49489/Intel-Atom-processor-D425-(512K-Cache-1_80-GHz)

I can of course set the TJMax in the settings which is great, but after restarting the program it loses the value I have set and goes back to 90°C

This is what I have in the RealTemp.ini:

TjMax0=100
TjMax1=
TjMax2=
TjMax3=

Is there a way to force the TJMax value to be set and saved even after program restarts? Or should it be doing that already?

Many thanks for any help you can provide.

 

[unclewebb]

RealTemp 3.69.1
http://www.mediafire.com/?4uixpjtezznuzkd

It should be saving the TJMax value correctly. Where do you have the RealTemp folder located? If you drag that folder to a location that you don't have read / write access to, Windows might be blocking RealTemp from writing to the RealTemp.ini file where it stores all of the settings. Try right clicking on that file, select Properties, click on the Security tab and make sure you have full read / write access to fix this problem. Let me know if this works.

 

[Dauntless]

Thanks for the quick response!

I put the RealTemp folder in: C:\Program Files\RealTemp

I just checked and the Administrators group has full control to the directory and files (I'm logged on as an administrator)

Also, it does save the other options I have turned on such as Start Minimized and Enable Logging

The same thing seems to be happening with the link you just gave

 

[unclewebb]

I have a laptop with a Core 2 Duo and TJMax is saved and restored correctly. I'll have a look at my code when I get a chance to see if I can figure out what's going on. It seems like TJMax is being saved correctly for you but is not being read from the INI file correctly. If I can't figure this out, I guess I will have to at least use the correct 100C TJMax value for your CPU. Can you post a CPU-Z screen shot of your CPU? The CPU name string will help me figure out what's going on. Thanks for the feedback.

You can upload image files to www.imageshack.us and then just post a link here.

 

[Dauntless]

I've just realised I stupidly copied the wrong version when trying out your new fix... (I extracted 3.69 instead of 3.69.1) Sorry about that, my desktop is cluttered

The 3.69.1 reads the correct TJMax straight away. Brilliant! Thank you very much!

I've double and tripple checked with version 3.69 and it's definitely reading the wrong value and not keeping the new value after a restart.

Incase you still need it I've attached the screenshot as requested

Here is what I had in the 3.60 ini file

WindowXY=557458180
IconTextColor0=0xFFFFFF
IconTextColor1=
IconTextColor2=
IconTextColor3=
TjMax0=70
TjMax1=100
TjMax2=100
TjMax3=100
Idle0=0.0
Idle1=
Idle2=
Idle3=
IconShow=1
LogFile=1
LogInterval=60
StartMinimized=1
TaskBar=0
FontNumber=
OnTop=0
TrayInfo=1
InfoWindow=
MiniMode=0
MinimizeOnClose=1
Spec=         Intel(R) Atom(TM) CPU D425   @ 1.80GHz
RivaFile=
Alarm=0
Alarm1=100
Alarm2=100
CSV=0
AlarmFile=
VIDInfo=0
ClockMod=8
GUIColor=0xF4F3EE
GUIMiniColor=0xF4F3EE
GUIText=0x000000
GUIMiniText=0x000000
GUINormHead=0x000000
GUIMiniHead=0x000000
NormHeading=0
MiniHeading=0
TMLoad=0
CustomGUI=0
CustomGUIMini=0
CPUID=0x106CA
Quality=0
ThreeDigit=1
GPU=0
Fahrenheit=0

 

[unclewebb]

Thanks for letting me know that RealTemp 3.69.1 is reading TJMax from the INI file correctly. I thought I remembered fixing that!

Posting screen shots and INI files makes my job a lot easier.

I'll go have a look at RealTemp and I'll make sure the next version is using the correct 100C TJMax for your Atom D425 by default.

 

[ndepriest]

Anyone know what is wrong with my computer. I have a cooler master V8 and applied Arctic Silver, temps should not be that high.

 

[unclewebb]

The i7-960 puts out a huge amount of heat but you're right, the temperatures shouldn't be that high. All you can do is pull the heatsink and try again. Have a good look at it and make sure it is perfectly flat. I would also pull the CPU and take a straight edge to it to make sure it is perfectly flat. Check it from side to side as well as from corner to corner. Doing this by a window can make it a little easier to see if there is daylight getting under the straight edge.

If so, your problem might be with the heat spreader over top of the CPU. When this is not flat then the heatsink doesn't make good contact with the CPU and you will get poor thermal transfer and high temperatures. Also make sure your case has good air flow. If that is OK then what type of thermal paste did you use and did you apply it according to the manufacturers directions?

The large difference between core temperatures of the center two cores is usually a good sign that something isn't quite right.