Printable View
I tested this on Vista Ultimate x64, ASUS Maximus Formula and a QX9650 cpu.
It does show 10 degrees less than Everest, but I believe Everest is likely closer to the true temps.
At full load (OCCT cpu test), I get about 52 degrees across the cores in Everest, 42 degrees in Realtemp. This is at 4 GHz with 1.42v vCore.
My ambient is 24 degrees.
Sweet program, but it only shows 4 cores on my dual harpertown.:shrug:
Working well here in Vista 64 also. Specs in sig. Thanks for the effort!!! :up:
I don't think the program was designed for dual procs.Quote:
Originally Posted by KaptainBlaZzed
If I find an easy way for you to read the two CPUs individually, I'll add that feature and send it your way for some beta testing. Now that everyone seems reasonably happy with the basic program, I'll have some time to start adding a few extra features. CoreTemp has had a two year head start in development but I'd like to think that RealTemp has done a pretty good job of catching up during its first two weeks.Quote:
Originally Posted by KaptainBlaZzed
@unclewebb - yeah dude, you did a great job with it. Will you consider a user defined tjmax option? It'll be nice for comparison purposes for those of us with "unknown" chips such as my B3 Q6600. I'd like to have the ability to play with it [tjmax] on-the-fly.
What are your thoughts?
SpeedFan gives you the option of setting an offset if you don't agree with its choice of TjMax. I'll try to include a similar feature. I'm pretty confident with TjMax=85C for your B3 Quad and TjMax=95C for the G0 Quads.
Cool dude. Might be easier to just allow direct changing of the veriable rather than introducing a 2nd correction factor (already have the -- - o + ++ ones).
RealTemp 2.11:
Maybe I should call this the GraySky edition!
For GraySky and anyone else that can't accept the TjMax I have chosen for their processor or even for users that just want to make some comparisons to my CoreTemp competition, then this is the version for you.
I've added a file to the distribution called RealTemp.ini
It's a simple text file that contains the following info but really only needs to contain a single character: 1, 2 or 3. By setting it to one of the following values here's what you can do:
0
---------------------------
TjMax adjustment feature
---------------------------
0 NO adjustment
1 + 5C adjustment
2 +10C adjustment
3 +15C adjustment
If you don't need this feature then you can delete the RealTemp.ini file or you can continue using version 2.1 which is exactly the same as this version except for this one feature.
This is a temporary beta version and the structure of the .ini file will likely change in the near future to accommodate additional features.
http://www.fileden.com/files/2008/3/...7/RealTemp.zip
Any chance of getting vista Sidebar support similar to Everest? :)
Quote:
Originally Posted by unclewebb
it appears it's the 2.1 version still when i downloaded a few minutes ago? or am i an idiot:)
Cool dude, thanks for the feature.Quote:
Originally Posted by unclewebb
Sounds like a great feature but don't hold your breath waiting for it.Quote:
Originally Posted by WFO
and while you've nowt else on, can you knock together a decent OS :rotf:Quote:
Originally Posted by WFO View Post
Any chance of getting vista Sidebar support similar to Everest?
Sounds like a great feature but don't hold your breath waiting for it.
http://i38.photobucket.com/albums/e1.../Dibujo-38.png
I need to add +15º.
Why? So you can have an inaccurate core temperature just like Core Temp is providing you with?Quote:
Originally Posted by Unrealer
RealTemp version 2.11 is perfect for you. Edit the RealTemp.ini file and set it to 3 and you will now have core temperatures just as inaccurate as what CoreTemp is providing you.
If you want an accurate core temperature reading then leave TjMax=85C for your processor and set the Idle Calibration to ++. You need to read the documentation for RealTemp located in post #1.
It's not the fault of all these temps reading programs, it's the fault of the sensors of both the chip and the MB. It's like it is just a afterthought of design. Half of them seem to be getting stuck anyway. Flying blind really.
Asus Probe Says Your CPU Is Freezing Your ass on 28 - 30 idle and 38 - 40 load. Bios says Your CPU is freezing your ass on 26 - 30 idle 37 - 41 load.Core Temp says your CPU is boiling ass of your Core 3 - 4 40 idle 50 load and cores 1 - 2 are getting warm on 36 idle 45 load ..
I'm more than confused
Quote:
Originally Posted by unclewebb
For Q6600 "G0"s the idle calibration should be 0 right ?
Every processor is different. I included the Idle Calibration feature so each user could calibrate this software to their specific processor. How to get yourself in the ballpark is outlined in post#1. If that is too much work then setting it to 0 is best.Quote:
Originally Posted by Demo
Quote:
Originally Posted by unclewebb
Understood, will try the 266x6 method when i can get another room thermometer as my current room thermometer is far from accuarate, its night time here and a bit cold but its reading 30c ( w/c feels like 22~24 )
I was still unsure about why RealTemp and CoreTemp are reading the TJmaxx differently (also causing a 15c change in temperatures), so I posted a thread at OCN for people to voice their opinions and talk about it. Here's the link, it's a pretty good read:
http://www.overclock.net/intel-cpus/...-realtemp.html
(If I'm not allowed to link to outside forums, please inform me so I can take it down.)
And another member just did a test to see where the cpu would actually throttle, and which program is correct. For now this test only applies to the e2xxx and e4xxx series that are the M0 stepping.
http://www.overclock.net/intel-cpus/...tjmax-dts.html
RealTemp sets TjMax=95C for the M0 stepping. For the L2 stepping processors it only uses TjMax=85C. Go have a look at the Intel docs. The L2 E4300 uses the same TCase temp rating as the other B2 stepping Conroe processors from that era. The B2 E6300, E6400, E6600 and E6700 are all TjMax=85C. It makes no sense that the E4300 would be significantly different.
CoreTemp set TjMax to 100C for the L2 E4300 and other L2 processors after it came out and people were getting below ambient idle temperatures. Raising TjMax for these processors was a mistake. The programmer of CoreTemp did not take into account the problems with the DTS reporting idle temperatures which effects all core processors. Have a good read of post#1.
If I can find an L2 processor locally tomorrow I might buy one and try to prove this once and for all.
Setting TjMax=85C and using a ++ idle temp correction factor will get you the most accurate temps for your L2 and will cure the below ambient readings that well cooled L2 processors report.
Edit: I noticed during testing that the PROCHOT# history bit first got set when DTS=2.
How come people report hitting below ambient when the tjmx is 85c?Quote:
Originally Posted by unclewebb
Thanks for responding by the way. I am really hoping to get all of this resolved so I can overclock some more. :D
You've been amazingly helpful unclewebb.
Hi Uncleweb.
I just finally read all the posts on the thread.
I like the work and effort you put within this tests.
I did some tests 2 days ago si I can helo you with the real Temps measured.
Sorry I couldnt post the tests before but I had a vacation weekend.
Now, Here is it:
Sorry for the pics, I was testing on a rainy day.
First, I took my UEIDT200 Thermometer and conected probe 1 to the CPU IHS and probe 2 just hanged on so you can see the ambient temps.
You can also compare ambient temps reported by my weather meter:
http://bandwidth.se/imgs/53/191/0.jpg
So, you see.... almost 25 Ambient.
Now CPU used is Core 2 E8400 so this can help with the tests.
The way I used the probe on the CPU was simple, just by one side touching the IHS of the CPU so I can measure the temp that way.
In fact, you did something similar by infrared termómeter, but I will meausre all the time CPU temps with my probe.
http://bandwidth.se/imgs/53/191/1.jpg
So I turned on the PC, wnet into the BIOS and set 1.125v CPU VID with all other voltages on stock VTT =1.1 and vNB on Auto as well as for SB.
As you can see, on BIOS CPU temp is 36 while actually is around 29 Celsius.
http://bandwidth.se/imgs/53/191/3.jpg
http://bandwidth.se/imgs/53/191/2.jpg
Now, for the tests, I used 3 Software:
1.- Core Temp (which is obviously the worst tool for Wolfdale it seems)
2.- Motherboard Monitor 5 (with FCG atachment so I can read DTS)
3.- Real Temp (by uncleweb)
Now, the screens I took will show the temps on idloe and Load. Tried SuperPi and Orthos on Both Cores.
Here CPU stock on idle temps:
http://bandwidth.se/thumbs/5902.jpg
While temp on sensor is about 27 degrees, Core Temp reports temps way to high and MBM reports temps below ambient.
Notice Real Temp is reportint 29 C but the calibration is - so, normally it would be around 31 Celsius.
If I put calibration on --:
http://bandwidth.se/thumbs/5898.jpg
Now:
http://bandwidth.se/thumbs/5907.jpg
While doing SuperPi. Thermal probe reports up to 29.6 Celsius.Core Temp again is way to wrong and if you see, Real Temp shows 35, and MBM shows 29 Celsius.
Seems like MBM is getting a good temp reading.
Let me clear that MBM as far as FCG told me, takes the Tjunction as 100º. While this doent works for idle, because it reports temps under ambient, seems to work on load.
Now, on orthos:
http://bandwidth.se/thumbs/5908.jpg
Notice Thermal probe is 32.4 Celsius.
Real Temp reports as 38 Celsius and again MBM reports 32 Celsius. So, it seems MBM gets good readings on Load.
Here on some minutes more after running:
http://bandwidth.se/thumbs/5901.jpg
You can see MBM says the exact same temp as Temp sensor.
At this point, Im starting to think MBM could work on load temps, but why is it so off when idling reporting always below ambient temps?
Now, I will set 3.8ghz with 1.32v on CPU.
Here the voltages set on BIOS:
http://bandwidth.se/imgs/53/191/10.jpg
Here on idle:
http://bandwidth.se/thumbs/5906.jpg
While Temp sensor shows 29 Celsius. MBM is below ambient, Core Temp is way too high and Real Temp shows 30 Celsius when using -- calibration.
While running SuperPi 1 core @ 3.8ghz 1.32v:
http://bandwidth.se/thumbs/5899.jpg
Temp probe shows 32.3 C.
While MBM is very close showing 32 C, real temp goes off up to 37 and I will not consider Core Temp which reads higher temps.
While running orthos @ 3.8ghz 1.32v:
http://bandwidth.se/thumbs/5904.jpg
Now, you see temp probe says 38.6 Celsius.
Now, neither of the temp programs report the temps acurately. Not even MBM which was working ok (or it seemed so) on Load temps.
While you can run the CPU passively, you can notice that the temp is really low for this CPUs. So, you can really take appart core temps readings and ignore them because I can even install OS with CPU without hetsink without problems.
Also, seemslike temps are lower than reported by any Temp program, and, for example, the MBM reports aparently acurate temps when load.
DIfference between MBM and Real Temp Tjunction is about 5 degrees.
- Also, MBM shows temps below ambient which are not possible when idling, and Real Temp gets good or at least, very close to real readings when idle.
- It seems that when adding voltage, all CPU programs show temp sclaling higher than it should be?
Well. I hope this can help you guys to make the best Temps reading program. I will make any tests requested if possible whenever this weekend, aganh, so dont hesitate to ask me some configs to test.
Kain