Real Temp - New temp program for Intel Core processors

[BeastNotro]

Umm ... I like the added color schemes you can choose what you want .. as for mini mode .. i checked it off .... but how u get to use mini mode as it not going to mini mode for me?

[unclewebb]

Ya gotta read the manual to get to Mini Mode. Try double left clicking on the user interface and be prepared. A second double left mouse click will bring you back. Windows Task Manager also has a similar feature accessed this way so it's not a totally foreign interface concept.

The other documented feature is Anchor Mode. If you double right mouse click on the user interface it will move RealTemp up to the corner of your screen. If you can think of a better position for it then move RealTemp to where you like it. Next, hold down the shift key while double right clicking on the user interface. This sets your new custom Anchor position so RealTemp will move to this spot whenever you do a double right click now.

[Ol'Baditude]

Oh my, that's sexy. Easy to use too. How sad is it to say that I never noticed that button at the top to switch the info from VID to CPU freq. to FSBxMulti. Lol.


And b4 anyone say anything about my temps, I have a 34Deg. ambient in here today. It's a scorcher with no AC. Booo.

[BeastNotro]

Unclewebb thats awesome i like it ... i like it alot...

Three Tumbs up for ya...

[unclewebb]

How sad is it to say that I never noticed that button at the top to switch the info from VID to CPU freq. to FSBxMulti. Lol.

Oh my gawd! I'm going to have to admit to RejZoR that my famous button is indeed a little obscure! Where's the sense of adventure guys? Click away on stuff! It's just a temp monitoring program after all.

Now do we need a background picture like Everest uses? RejZoR, sharpen your pencil. Any ideas for a background image?

[Lu523]

I am going to give the beta a try.

[Gitsum]

[unclewebb]

I am going to give the beta a try.

Don't worry, beta is not like a dirty 4 letter word.

Even the betas of RealTemp run pretty smooth and they don't gobble too many resources so don't be afraid to take it for a good run.

[RejZoR]

Oh my gawd! I'm going to have to admit to RejZoR that my famous button is indeed a little obscure! Where's the sense of adventure guys? Click away on stuff! It's just a temp monitoring program after all.

Now do we need a background picture like Everest uses? RejZoR, sharpen your pencil. Any ideas for a background image?

As i said before, that button has to tell the user what it does...
Here, i made a reconstruction how it should really look.



It just looks better because of the slightly different CPU name and info alignment, plus the button tells users exactly what it does. It shows more info. Oh and don't forget to make it the same lenght as those 4 buttons below, i've made a reconstruction and noticed that i made it slightly longer than those below...

Regarding background, i don't think it's worth investing time into it and in the end it'll just make reading values in window harder to read.
Everest is already using it's custom skining for entire window which i think looks ugly (some crippled hybrid between custom skin and XP theme). Yuck.
Though the internals of the window look pleasing because of the calm blue gradients used.

Anyway, stick to the plain interface. Trust me when i tell you this, many programs tried the skinning and failed miserably. So just don't go that way.
All RealTemp really needed was ability to use Windows visual styles (themes).
This way it always blends perfectly into the environment regardless of what Windows theme users are using.

[JohnZS]

I tried the new BETA...and I liked it
Once again Unclewebb you do a tremendous job, keep up the good work. I am curious to see what Intel have to say at IDF re: the TJ Max.
By the way, do you think I have 2 defective sensors?
My ambient temperature is ~20-25C (depending on how hot the weather is).
Here is an attached screenshot of successful completion of IntelBurnTest

What do you make of the temps for cores 3 and 4?
Thanks
John

[unclewebb]

The problem is that there is not enough room for what you suggest.



For your processor there is enough room but if you have an Intel Quad QX6700 ES processor and you are running it at 333.33 x 10.0 MHz then all of that isn't going to fit. With your full size button idea the only way I could do that is to combine the two boxes at the top into one box so a user could see half as much information. The present version shows processor name and MHz which tells me what I'd like to know in a screen shot. I'm open to new ideas but I'll have to wait for some more user feedback on this topic before making any changes. I could also scrap all of the additional info like Intel, Quad, Extreme and just stick to the model number like QX6700. That would work.

JohnZS: It is a good thing if you have two sensors that can display numbers that low. Most sensors get stuck before they can ever get that low. Read the RealTemp documentation about Calibration and try doing that procedure. With some minor adjustments, you should be able to get some very accurate core temperatures out of that processor.

In a couple of more days Intel should be telling us a lot more about these sensors and what programmers can do to get accurate temperature information out of them. The RealTemp calibration method was the easiest thing I could come up with but if a new, Intel approved, method comes out then I plan to update RealTemp to reflect that. If they come out and say TjMax=105C for all 45nm desktop processors then I'll probably ignore that because it doesn't agree with any of the testing I've done.

[RejZoR]

Well, you could certanly scrap "Intel" as you can't use RealTemp on anything else anyway... There you go, you just gained 6 characters of space.

[BeastNotro]

Just make the original button a black or bright red color so that ppl can see it better as it is blending in with the stock background... i have to admit i never even knew it was there until it was pointed out in here.... seriously . i didn't and i have been using realtemp since version 2.60 or just before that.

[Lu523]

I like the way it looks. Thanks for all the time you have spent on it.

[JohnZS]

JohnZS: It is a good thing if you have two sensors that can display numbers that low. Most sensors get stuck before they can ever get that low. Read the RealTemp documentation about Calibration and try doing that procedure. With some minor adjustments, you should be able to get some very accurate core temperatures out of that processor.

In a couple of more days Intel should be telling us a lot more about these sensors and what programmers can do to get accurate temperature information out of them. The RealTemp calibration method was the easiest thing I could come up with but if a new, Intel approved, method comes out then I plan to update RealTemp to reflect that. If they come out and say TjMax=105C for all 45nm desktop processors then I'll probably ignore that because it doesn't agree with any of the testing I've done.

Thanks Unclewebb
Your guide was informative and easy to follow, quite interesting to. I could not use the lowest voltage setting in my BIOS as for some reason I kept getting "CPU OVERVOLT" errors on POST.
I ended up doing calibration @ 1.00V (0.982V in CPU-Z) 6.0*266Mhz FSB.

I then rebooted @ 1.20V 8x400Mhz FSB and now have temps which are more or less the same

Keep up the good work...all eyes on IDF.

What if Intel say Tjmax IS...neither 105 or 95?!?!
John

[unclewebb]

JohnZS: The most Extreme thing about your Quad seems to be the sensors. My Q6600 isn't great but the Distance to TjMax which represents the raw data coming from the sensors only varies by about 4 or 5 degrees at idle from core to core. Your center two cores show a difference of 14!

All I can say is good luck to Intel to try and come up with a simple, user friendly formula to cover the wide variety of sensors that I've seen. Trying to cover a difference of 14 is stretching RealTemp's capabilities for covering up this mess.

If you're bored John could you try running a test with the calibration factors you're using? Run at whatever your normal MHz / core voltage is. Turn on the RT logging feature and set it to 1 second. Run 1 minute idle, followed by about 2 minutes of Prime95 small FFTs on all cores and then back to idle for a minute. E-mail or PM me the log file when you're done.

It is normal during this test for core0 / core1 to closely track each other and then for core2 / core3 to closely track each other. Core2/core3 being about 5C less while running Prime compared to core0/core1 is also fairly normal. I'm just curious to see how all these cores track each other during transition from idle to full load and back again.

Prime95 v25.6 32 bit
Prime95 v25.6 64 bit

One theory I have is that when a processor has badly mismatched sensors, Windows XP or maybe the processor itself will schedule a bigger work load on the core that reports the lowest temperature. When uncorrected, core2 reports that it is running so cool that it will get to do some extra work. To test for this run Prime95 and for Number of torture test threads to run enter 2. This will run 2 instances of Prime but they will be constantly spread to your 4 cores so no one core gets too hot. Do a CTRL+ALT+DELETE and bring up the Task Manager and you should get to see what core is getting the biggest chunk of work. I'm not sure if that's important but I found it kind of interesting.



My core2 reads highest Distance to TjMax so it gets to work the hardest. The load seems to be balanced for both sets of cores. Core0/core1 share half the load and core2/core3 share the other half.

[gymenii]

Just an idea

[JohnZS]

JohnZS: The most Extreme thing about your Quad seems to be the sensors. My Q6600 isn't great but the Distance to TjMax which represents the raw data coming from the sensors only varies by about 4 or 5 degrees at idle from core to core. Your center two cores show a difference of 14!

All I can say is good luck to Intel to try and come up with a simple, user friendly formula to cover the wide variety of sensors that I've seen. Trying to cover a difference of 14 is stretching RealTemp's capabilities for covering up this mess.

If you're bored John could you try running a test with the calibration factors you're using? Run at whatever your normal MHz / core voltage is. Turn on the RT logging feature and set it to 1 second. Run 1 minute idle, followed by about 2 minutes of Prime95 small FFTs on all cores and then back to idle for a minute. E-mail or PM me the log file when you're done.

It is normal during this test for core0 / core1 to closely track each other and then for core2 / core3 to closely track each other. Core2/core3 being about 5C less while running Prime compared to core0/core1 is also fairly normal. I'm just curious to see how all these cores track each other during transition from idle to full load and back again.

Prime95 v25.6 32 bit
Prime95 v25.6 64 bit

One theory I have is that when a processor has badly mismatched sensors, Windows XP or maybe the processor itself will schedule a bigger work load on the core that reports the lowest temperature. When uncorrected, core2 reports that it is running so cool that it will get to do some extra work. To test for this run Prime95 and for Number of torture test threads to run enter 2. This will run 2 instances of Prime but they will be constantly spread to your 4 cores so no one core gets too hot. Do a CTRL+ALT+DELETE and bring up the Task Manager and you should get to see what core is getting the biggest chunk of work. I'm not sure if that's important but I found it kind of interesting.



My core2 reads highest Distance to TjMax so it gets to work the hardest. The load seems to be balanced for both sets of cores. Core0/core1 share half the load and core2/core3 share the other half.

I guess that is why they call them Extreme Editions
You will have a PM coming your way shortly, I was idle for ~minute then maxed out all 4 cores 100% in Prime95 64bit Small FFT's for around 2mins.

You will probably understand the results more than myself, but it looks like my calibration is useless at over 50C, all cores are within 1-3C of each other until 50C is met, then a "divide" between cores 0+1 and 2+3 is visible (0+1 running hotter than 2+3), yet when I take the stress off the CPU and it cools down, at around 49C the numbers all level off again.

Now get this, this is my 2nd QX9650, I had to RMA my 1st one which I purchased in December 07 because Core 3 kept Prochotting and causing the PC to freeze under high load. AND the sensors were worse.

Needless to say my replacement was stable and still had dodgy sensors but stability is all that matters really.

Anyway PM enroute. Thanks for your help and I have seen your theory in practice. With wPRIME one of my cores lags behind the rest, although wPRime calls it Core3, I belive it is Core1 as that one runs the hottest at calibrated settings...now if only intel released a calibration tool for the microcode or something....maybe integrate with RealTemp

John

[unclewebb]

This is just a quick version to show what RealTemp could look like.



It's available in a separate folder from the normal beta stuff:

http://www.fileden.com/files/2008/3/...lTempBeta2.zip

I dumped all of the extras so just the model is displayed. The extras like Intel, Quad, Extreme, Pentium and ES aren't really that important. There's enough room in the middle for the extra info and the button is full size, easy to understand and use.

I kind of like your idea gymenii. I'll see what feedback I get about this. I figure if two different people in the same forum on the same page didn't understand that my small button existed then I need to do something to get people's attention. Of course, they were both from Canada! It might just be something in the water. (No need to flame me, I live here too.)

My competition has a new version out:
Core Temp 0.99.2

It has a few features like a Vista Widget, graphing abilities and G15 keyboard support that RealTemp doesn't.

[Lu523]

I kind of like your idea gymenii. I'll see what feedback I get about this. I figure if two different people in the same forum on the same page didn't understand that my small button existed then I need to do something to get people's attention.

I did not know it was there until I did some reading here. I think I like the cleaner look without the info in the button though.

[BeastNotro]

Unclewebb

You need to do this or i will be using core temp..

You need to add the G15 implementation for the LCD, right now i can see my core temps and CPU load on my G15 keyboard it is v2 of the G15 model.

[unclewebb]

I'll put G15 support on the things to do list. Maybe next month. Personally I also like the small info button. Labelling it with the international i symbol might bring it to people's attention better. Final decision on what to do will be based on user feedback.

[BeastNotro]

Yes i like the little i = info button idea and thks for the G15 looking into it for me. Its just a cool feature as when playing games and such i can see my temps on my LCD and i like real-temp better then core-temp .. so get on it faster if you can....

thks take u time as what i do for now is run core-temp when playing games so i can see temps and cpu load on my g15 lcd screen.

Also for the info button .. add text to state press here 1-6 times for all info abut CPU.. as one click . some ppl are not bright ya know.....

Oh and core-temp also shows on lcd screen my ghz and what i am at 8x450 and Vid too . plus it says Intel Core 2 Quad Q6600 (kentsfield) on LCD to... that cool features .. dude

Maybe i take a pic and show ya it. if you not have a G15 keyboard.

Edit:

Here is picture of it on my LCD screen.

[loonym]

I like the small button too, I personally don't care how or if you label it. The ones that need labels and mouse overs are the same ones that will be here asking 'how do I calibrate it?' or 'what do I believe?'

[Ol'Baditude]

I'll put G15 support on the things to do list. Maybe next month. Personally I also like the small info button. Labelling it with the international i symbol might bring it to people's attention better. Final decision on what to do will be based on user feedback.

I couldn't care less TBH. But if you were to do something about it, please keep the button the same size and maybe put a little "i" in it or something. That would do the trick and remain unobtrusive at the same time.