Real Temp - New temp program for Intel Core processors

[unclewebb]

Almost 600 downloads so far and no complaints.

Thanks to all the beta testers here at XS for helping to get the bugs out.

If you find any problems let me know so I can get it fixed up.

[theonlybabyface]

When using the "Gamer Mode" how can I prevent the screen from continously flashing? Also, any thoughts on a Vista gadget similar to the one for CoreTemp?

Great program btw...

[unclewebb]

theonlybabyface: Unfortunately, there's presently no way to prevent flickering in Gamer Mode. It will vary depending on the frame rate of your game and the screen background color. It's just something I'm experimenting with so I thought I'd share it and get some feedback. At the moment, if you can get any information at all out of Gamer Mode then consider yourself lucky!

I'll be looking into using RivaTuner to implement this feature correctly. From the feedback so far, I think it would be worth some further development. A Vista gadget and G15 keyboard support is still somewhere off in the future.

[loonym]

Nice app any Idea of which is right I am running Vista 64bit. I know it is reading the wrong MHz on my Q6700 but I am not sure about the rest.

From the reading I've done, it seems this type of question is very difficult to answer without knowing the precise details of the setup. The hardware, how it's setup, cooling, airflow, ambient temp, all come into play. One or neither of the examples in the screenshot may be right. That's the beauty of realtemp's highly configureable interface. You can make it as right as you want with the settings according to the facts as they apply to your particular rig.

[SawyerX]

My system is:
Q9300, ASUS P5K PRO, Mugen cooler+120mm. Closed Case with AIR filters all arund and 2 additional 120mm 1000rmph fans blowing into. One at front and one at back into the Mugen. So theoretically there should be more fresh air inside then in a solid box.
Ambient temperature meassured with a thermometer is 22.6C. Airconditioner running right now.



Thats the picture. I also did install ASUS probe 2 to check CPU temp and it reports same temp as HardwareMonitor CPUTIN 37C. Whats reporting this exactly? Are there 5 thermal sensors? Like 1 in each core and 1 on the base or something?

Also what can you tell me about this temperatures, because I somehow think they are to hight for a Mugen cooler, but dont really know.
I did read your manual on the webpage and understand that nothing can really read idle temperatures good right now. SO is my TJ max at last ok or should I worry?

I also dont fully undesrstand what you are saying with non linear reading. Do you mean the temperatures go wrong with time like the longer its running or as they build up?
I have noticed if I install ASUS AIgear3.0 and even if it downclock my CPU to 0.96Vcore and like 1799Mhz. The actual CPUIN temeprature goes up. I can not understand why itsgoing up insead of down other then the sensor goes more wrong as time passes by. Calibration sensorreports ok like 3-4-3-3

[Grandpa]

unclewebb

Just noticed your PM 2.6 is reading the Mhz correctley now, nice job

[loonym]

Nice clock on that kent Grandpa

[Grandpa]

Thanks Loonym, I have had her stable at 4.1Ghz but way too much heat and Voltage, it took 1.85V to get it stable. I am very happy with the current clock considering I am running 8GB of Mushkin XP2-8500 (996599) at 1108Mhz 5-5-5-12-2T @ 2.1V

unclewebb

I think I have found a bug, if I go to settings and put a check mark in the MHz box it will read the MHz wrong. If i uncheck it, it is correct.

[unclewebb]

SawyerX: Motherboard based CPU temperature readings are rarely if ever 100% accurate but when they appear to be close, users assume that they are 100% accurate. A good example is your reading of 17C for AUXTIN. That's below your room temperature which is impossible so it's wrong. Knowing that this sensor is wrong casts doubt on your CPUTIN reading even though what it's showing is probably very close.

As nice as air-conditioners are, they can also cause temperature gradients within a room. I'm lucky that I have access to a cool, consistent basement year round.

I'll assume that your room temperature of 22.6C was measured close to your computer and that it is reasonably accurate. I find I get more consistent results with the case open. How much thermal paste, how you applied it and whether it has reached its maximum effectiveness also plays a part in this. I just want you to be aware that there are a lot of factors that influence the accuracy of your readings as well as the accuracy of my test.

To start with, the core temp readings of 50C to 52C as reported by CPUID Hardware Monitor and CoreTemp are wrong. There is no possible way that during this test that your core temperature is almost 30C above your room temperature. RealTemp uses TjMax=95C for your processor which at least gets your temps closer to reality. The minimum that RealTemp reports is 39C but I think that is still too high and I would recommend a negative calibration factor to reduce the reported temps further.

When my E8400 reports 40C with no calibration factors, I measured 35C with the IR thermometer. Given your room temperature, and a 101 other factors, I would try using these calibration factors:

( -1.9 , -3.0 , -1.9 , -1.9 )

This will result in minimum reported core temperatures at idle of 34C for all 4 cores which is very reasonable considering your room temperature and other details you've provided. My E8400 with TjMax=95C and a -2.0 calibration on both cores provides very accurate temperatures and I think your 45nm Quad with these recommended settings will also be very accurate.

Will it be perfect? Probably not but without buying an IR thermometer it's probably about as close as you're going to get. If you fully trust the CPUTIN reading and think a core temp reading of 35C or 36C is more realistic then adjust the above factors a little closer to zero to reduce the amount of correction. On a Quad, I try to balance the idle temps across all 4 cores. At full load there is typically a difference between core0/core1 and core2/core3.

Grandpa: I think I have found a bug, if I go to settings and put a check mark in the MHz box it will read the MHz wrong. If i uncheck it, it is correct.

Thanks for the feedback. The CPU MHz option works 100% on my board and is compatible with SetFSB and provides MHz readings equal to what CPU-Z reports. Unfortunately, on some motherboards, this option over reports the MHz. Not checking this box uses a different method to calculate MHz which should work on all motherboards but this method is not compatible with SetFSB. I guess you'll have to use CPU-Z if you are using SetFSB.

Anyone with this issue can PM me what CPU-Z reports and what RealTemp reports with CPU MHz enabled at a few different MHz settings and I'll try to come up with a fix. Include what brand and type of motherboard you're using.

[SawyerX]

Well allright. Gonna do the test again. Now I will downclock everything I know of. Open the box. Install the case thermal sensor and compare it to my other 2 room thermometers.

I will turn airconditioner off also.

[SawyerX]

Oh here is my second test.

1. A picture of how it looks inside. http://freeweb.siol.net/denisbr5/31052008066.jpg

2. Test done on Ambient 23.7 close to case. Later I did put the case thermal sensor close to cpu and it was 27.7C. Case was still open all around like on the picture and standing straight.
http://freeweb.siol.net/denisbr5/ambient.jpg

3. Aircodition with Aimbient temp 21.3C with thermometers and sensors close but outside the box. When I did place it close to cpu cooler inside with case still open it was showing 25C.
http://freeweb.siol.net//denisbr5/aircondition.jpg

So the difference is around 4C when thermal sensor is outside open box or inside open box.

4. Case is now closed and put in place with 30.3C, . Ambient around 22.1C. All 4 cores on 100%. CPU is still downclocked.
http://freeweb.siol.net/denisbr5/load.jpg

5. Default clock but speedstep, Idle. Vcore lower then default but seems stable.
Ambient 22.4C, Case temp 30.1C http://freeweb.siol.net/denisbr5/defaultclokcilde.jpg

6. Default clock, Load 100%. Ambient 22.4C, Case temp 32.4C
http://freeweb.siol.net/denisbr5/defaultclockload.jpg
PANIC TIME.

[Mk]

Real Temp caused BSOD when i was sleeping


I love real temp hope you can fix this I'm sending the file's windows kept for error reporting


http://rapidshare.com/files/11906074...files.zip.html

[Grandpa]

unclewebb

I noticed when running 3DMark06 and RealTemp 2.6 that when using gaming mode that 3DMark06 score takes quite a hit,(about -1000 points). I tested several times with the same results, if I turn gaming mode off there is no problem. Most of the hit is taken in the SM3.0 test.

[msgclb]

Real Temp caused BSOD when i was sleeping
...

Don't go to sleep!

What is PCA2?

I ask because a Google search for 'Vista PCA2' didn't give me a good answer. A lot of different applications report problems with this event but so far you're the first one that I could find that associated it with Real Temp.

I installed Real Temp v2.60 on my Vista x64 system and I can report that so far it is working. Maybe because I haven't tried sleeping!

I'll report back if Real Temp gives me a BSOD tonight.

It is also installed on my Vista x32 system and it has never caused a BSOD.

[unclewebb]

C.Ron7aldo: Can you explain your problem a little better. A BSOD is when the screen turns blue and the entire computer locks up and you have to push the Reset or ON/OFF button to continue. Did that happen to you? Did the error report show up after you restarted?

I tried to go to Rapidshare to download the files but I had to give up because I wasn't successful at counting cats. Rapidshare has gone to hell in my opinion. The FileDen doesn't use crap like that.

In the About... box of RealTemp there is an e-mail address you can use to send bug reports or whatever.
Real_Temp@yahoo.ca

I hate posting e-mail addresses in forums because I'm already getting a pile of offers to claim my million dollar prizes from Nigeria!

Last month was about adding features. I'm going to take a break from adding anything new until I'm sure that RealTemp is bullet proof. I've been constantly adding and changing things that I haven't had time to do any long term testing but that will happen in both XP and Vista.

When reporting errors, try to let me know what specific features you were using so I have a better chance of tracking things down. A screen shot of the Settings you were using will help. Also include your opinion of how stable your computer is and any other info you can think of.

SawyerX: I do calibration testing with an open case. The reading from CPUTIN seems very close to the truth in my opinion and during the idle test I would use calibration factors for RealTemp so that it reports temps the same as this sensor. Your calibration settings will be less than I recommended before because you have a large temp gradient depending on whether your case is open or closed.

In this picture: http://freeweb.siol.net//denisbr5/aircondition.jpg
you say the air temp is about 25C and CPUTIN is reporting 32C. With a Quad core a difference of 7C is typical during this test. I'd calibrate RealTemp to report 32C. Intel has shown during testing that the difference between a core temperature reading and a Tcase temperature reading at idle is less than 0.5C so CPUTIN and core temp readings being the same is reasonable. Try using the Reset button in RealTemp during each test so I can see exactly what the Min and Max temps were during your test.

Grandpa: That's motivation enough to get this feature done properly. My score only dropped about 10 points but I'm using an older graphics card so the hit wasn't as bad.

[msgclb]

There is a feature in Vista called 'Problem Reports and Solutions'. This was a report from v2.58 where Real Temp stopped responding during XS Bench.



This is a true BSOD. (not related to Real Temp)



The question I have is would these reports be of any good in troubleshooting Real Temp problems?

[unclewebb]

Thanks for clearing that up msgclb. Version 2.58 had a couple of pointers pointing to nowhere and the program choked during the XS Bench. Saving data to a NULL pointer is not a good thing. I think that beta version was alive for about 7 hours while I was snoozing one night but was quickly replaced the next morning when I saw that I had screwed up.

The Windows generated error reports probably aren't going to tell me anything.

If RealTemp problems are more wide spread or repeatable then it'll be easier to find and correct them. I'm using XP and haven't had any problems since version 2.58 was fixed but I'll be going through the program, line by line, to look for anything that needs improvement or could be causing trouble.

[Massive]

sorry for the noob question.
I still confuse why is there several temp reading on different program, why just manually test the real temp at the core processor, or at any place that we're all agreed that it is the most reliable place to show the "REAL" temp, and then compare it which program come closest with the manual reading.

I'm overclocking my proc right now, and currently using core temp right, and I'm very careful when it comes to temp so I won't fry my cpu. Abit default EQ is showing a lot more higher temp then core temp, so I'm using core temp since I've read it is more reliable. And From what I've read too, core temp show higher value then real temp, I'd felt more safe knowing the higher value so I won't overclocked it too much.
Or Is there anyway to find that real temp is reading the "true" temp of my proc?
Btw I use Abit IP35.

[unclewebb]

sorry for the noob question.
I still confuse why is there several temp reading on different program, why just manually test the real temp at the core processor, or at any place that we're all agreed that it is the most reliable place to show the "REAL" temp, and then compare it which program come closest with the manual reading.

If you took the time to read the RealTemp docs then you might discover that is kind of what RealTemp is all about. I wasn't too impressed with CoreTemp reporting my E8400 15C to 18C too high at idle but if big, inaccurate temp numbers make you feel safe then stick with CoreTemp.
-------------------------------------------------------------------------

I've started rewriting RealTemp to make it more efficient and to cause less stress on the system resources. I just completed a 10 hour test with RT2.60 and the new RT2.61. Neither version crashed and when minimized to the System Tray, the improvement in efficiency is obvious.



Beta testing will resume shortly. I really need someone that is having trouble with the CPU MHz option to PM me some numbers for CPU-Z vs RealTemp at a few different frequencies while using SetFSB so I can finally try and correct this issue.

[Grandpa]

unclewebb

I tried to send you a PM , but I kept getting a data base error.
Let me know if you need anything else.

Setting CPUZ RT/WO RT/With Diffrence
490 X 7 3430 3430 3485.7 55.7
462 X 8 3696 3696 3756 60
400 X 9 3599.9 3599.9 3658.3 58.4
411 X 9 3699.2 3699.2 3759.2 60
380 X 10 3799.8 3799.8 3861.5 61.7
370 X 10 3699.8 3699.8 3759.8 60

[unclewebb]

Thanks Grandpa. The data base error usually means the PM system is down for maintenance or similar reason.

The numbers you've posted will help me come up with a patch.

[loonym]

unclewebb, I was a little surprised when I swapped my x3360 from an x48 to a 780i and my vid was different. Maybe I'm confused as to how vid actually works. Is what I experienced normal? Although thinking it over, while unexpected, it is 'normal' since it's what happened. Here's screens, X48 and 780i, same x3360 in both.

[CERO]

So Tjmax for the Q6600 is 95c huh. I always used coretemp, and I havent that much around on XS , but recently i see allot of peeps using real temp, how did u find out that the Tjmax is 95c unclewebb? So it means if i trust this program, that Coretemp was reading all the time 10c higher then the "real cpu temp"?
And what is this Max en min temperature display? If the Tjmax is assued at 95c, does it still apply that the max temp for the CPU is 71c, according to Intel specs?

[loonym]

max/min are simply the highest and lowest recorded temps during the time realtemp is running. I think 71c intel limit is tcase max which is quite different than actual core temperature.

[unclewebb]

loonym: You could try running CPU-Z and in the cpuz.ini file set

SENSOR=OFF

to see if CPU-Z reports the same VID as RealTemp.

During testing, RealTemp and CPU-Z always reported the same thing for VID. To be honest, the Intel docs I read were not clear when it came to the correct way to read VID. I assumed that CPU-Z was getting this right and tried to follow along. You might also want to try CoreTemp which I think recently changed its VID calculation for the 45nm chips.

I don't know if these CPUs have a VID number stored in the chip that is totally independent of the motherboard it is installed on. If anyone finds in the Intel documentation how to read VID information then post it here and I will compare that to what I'm doing to see if I screwed up.

I know on my wife's Dell laptop, T7200, that RealTemp and CPU-Z report two different voltages for VID depending on whether it is in CxE power saving mode or not.

CERO: My choice of TjMax=95C for a Q6600 - G0 was based on testing of my Q6600 with an IR thermometer.

The previous assumption by every other program that TjMax=100C seems to be based on that is what Intel TAT uses even though TAT was specifically designed for testing mobile core processors and was never updated or designed for use with Core desktop processors.

If you trust TAT more than you trust my testing then it is easy enough to go into RealTemp and set TjMax=100C.

Edit: Here's some info from my original test:
http://www.xtremesystems.org/forums/...&postcount=568

It was a brand new processor that I was unfamiliar with so I didn't do a thorough test but hopefully I will have some time later this week to re-test my Q6600. After using this processor for a couple of months, the temperature sensors on core0 and core1 seem to be the most reliable and give consistent / balanced results.