Real Temp - New temp program for Intel Core processors

[unclewebb]

Any chance of making the temps display in the system tray in a future release?

That's near the top of my things to do list. I just wanted to see how users would react to this program and so far so good. I don't mind doing some work if users are going to use Real Temp.

I saw a XS benchmark score posted of 534 when eva2000 was running at 1604 MHz. 1000 points is what a typical 3000 MHz E8400 scores so this bench says you are probably running at about 534 / 1000 X 3000 MHz = 1602 MHz. I told you that bench was accurate and scaled very linearly with CPU MHZ.

[Infa]

like this program very well

like someone mentioned, tray minimize would be sweet

and when i close the program, and start it again, i have to put back the calibration, its not saved when i close the prog. and i dont get the log to work, or does it not save in the same catalog? otherwise its working fiiine

[unclewebb]

There is not yet a save feature for the Calibration setting but that's easy enough to do.

The logging feature works and should save RealTempLog.txt into the same folder where RealTemp is located. Start up RealTemp, turn on the logging feature, wait 30 seconds, close the program and you should see the log file sitting beside the RealTemp program.

[[XC] 4X4N]

Tray minimize would be perfect. Here is idle and load on a Q6600. Great program

[Slay0r]

Nice program =) some readings with an e4300

Calibration @ 0

load


idle


Calibration @ ++

load


idle


I have given up trying to find out actual temps for this chip a while ago, so there, have this reply just as a FYI

p.s. e4300 is wrongly recognised as xeon by realtemp .

[unclewebb]

Reads sub-zero? Would be wicked sweet if it did

How far below zero are you going? The Core processors store the DTS temperature in 7 bits of data which gives a range of 0..127.

Temp = 95C - 127C = -32C

The liquid nitrogen guys will need to find another program!

There is a controversy of whether the L2 E4300 has a TjMax of 85C or 100C. I'm betting that the early E4300 had a TjMax=85C which is why my program reads 15C lower than others. Once I get my hands on one or a similar L2 processor I'll do some proper testing and hopefully get this cleared up once and for all. Core 2 Duos don't run very stable at a core temperature of 85C when you are overclocking and overvolting so that's why I'm betting that 70C is actually closer to the truth.

I'll get the Real Temp Xeon bug cleared up by tomorrow.

[idiotekniQues]

what if you dont know C++

im gonna cry!

[Kilyin]

what if you dont know C++

im gonna cry!

No need to know C++, just follow the installation instructions and run the program.

[Kenski]

Thanks for your hard work,
Just tried it out and seems to read closer to what I expected temps to be,
running a QX9650 @ 4Ghz (10x 400) watercooled and now seeing low '50's under load,
In the next few days I'll try a half speed test and see if the results match up, with an unlocked multi I should be able to run 12 x 333, and then 6 x 333 and compare the temps reported, that should show a 50% reduction in delta temps if the full load readings I'm getting are accurate.

[unclewebb]

Kenski: Does RealTemp read the QX9650 cpu string properly or does it screw it up and call it a X9650? I think I screwed up when I tried to add on detection for the Celeron Core based processors which only have 3 digit model numbers.

Slay0r: Can you try my test where you drop the multi to 6.0, on an E4300 you can drop the FSB to 200 MHz and drop your core voltage down to about 1.10 volts. Boot up and run your CPU fan as high as it'll go. The reason TjMax got boosted to 100C for the E4300 was because during this test programs were reporting idle temperatures below ambient. I think with my program and an 85C TjMax you will be able to set the Idle Calibration to ++ to correct this problem without screwing up your full load temps.

idiotekniQues: what if you dont know C++

No worries. There won't be a test or anything like that. If you know how to double click with the mouse you're good to go surfing and you can check out your temperatures as well. I just thought I would include the most important line of C++ code that this program uses so any other programmer can write their own program if they like.

[Kenski]

Kenski: Does RealTemp read the QX9650 cpu string properly or does it screw it up and call it a X9650? I think I screwed up when I tried to add on detection for the Celeron Core based processors which only have 3 digit model numbers.

Slay0r: Can you try my test where you drop the multi to 6.0, on an E4300 you can drop the FSB to 200 MHz and drop your core voltage down to about 1.10 volts. Boot up and run your CPU fan as high as it'll go. The reason TjMax got boosted to 100C for the E4300 was because during this test programs were reporting idle temperatures below ambient. I think with my program and an 85C TjMax you will be able to set the Idle Calibration to ++ to correct this problem without screwing up your full load temps.

I'm seeing Intel X9650 and Cpuid= 0x10676.

[Andrewv]

for my E2160.
The real TJ i think is 85!!

Coretemp shows 100

[tjh566]

Still reads different that uguru which I believe is correct with my phase.

While uguru says 15*c your program is reading 20+ and when I take off load and go subzero on the core it reads 4*c, whereas uguru just errors out with I go subzero.
but it is alot closer than coretemp is, coretemp was saying I was going into the 30's

[Infa]

uncleweb, cant get the log to work, it wont put out a file in the same folder..
and i have turned it on.

maybe i have to run it as admin?

[Slay0r]

Ok i'll get that sorted after dinner unclewebb.

By the way the fan RPM readings tend to go tits up on Hwmonitor, my cpu fan is the one named "Chassis 3" in Everest and its @ full incase you were wondering.

[unclewebb]

My testing showed that the DTS data at low temperatures simply isn't accurate.

My Idle Calibration can help out the situation but it's just a guess.

tjh566: Post some screen shots so I can have a look.

[unclewebb]

uncleweb, cant get the log to work, it wont put out a file in the same folder..
and i have turned it on.

maybe i have to run it as admin?

I think that depends on where you have the program sitting. If the actual program is sitting on your Desktop I think it should work but I'll check.

I just tried it on a Limited XP Account. With the program sitting on the desktop it was able to output the log file with no problems. Notepad screws up the formatting but it looked correct in Wordpad.

[JustusIV]

Can i do anything about my temps being stuck? Both cores seem to be "Stuck" one at 46C and the other at 40C. it is rather frustrating.
JustusIV

[unclewebb]

JustusIV: Does my program at least detect your two stuck cores as BAD? There isn't anything you can do. You can't get temperature data from a sensor that isn't reporting it correctly. Include a screen shot if you can.

[Extelleron]

Will this program work under Vista 64? Doesn't seem like it does now but will it in the future?

Just wondering as I'm debating Vista 64 vs 32 for my next build.

[unclewebb]

Will this program work under Vista 64? Doesn't seem like it does now...

I don't think it will work with Vista 64 without an updated driver but can anyone confirm this?

[Xvys]

Here is a screenie of my Test Sensor. My room temp is at least 22C-23C, so if core1 sensor is correct, my cores are running at ambient temp with core readings of 28C/22C (assuming the 28C is stuck and should be 22C like the unstuck core)? The 41C/43C max temps are after Prime95 session. I know my cooler is good, but.... I have the TK Ultra120 Ext fan at it's lowest setting (Antec Tri-Cool @ 1000 rpm). With this same cooler my E6300, with this same P5K-D, had idle core temps of 55C @ 3.45GHz with 1.375v vCore (with fan at 2650 rpm). I had thought my CoreTemp idle temp of 34C seemed about right with the E8400?

[RPGWiZaRD]

Mine's a bit weird still IMO.

[JustusIV]

JustusIV: Does my program at least detect your two stuck cores as BAD? There isn't anything you can do. You can't get temperature data from a sensor that isn't reporting it correctly. Include a screen shot if you can.

Here is what is going on. Realtemp says core 0 is bad but core 1 is ok. Is a movement of 7 somewhat normal? seems like a difference between load and idle should be more then that?

[ken]

my load temp using realtime and coretemp are exactly the same [no calibration]