Page 110 of 180 FirstFirst ... 1060100107108109110111112113120160 ... LastLast
Results 2,726 to 2,750 of 4486

Thread: Real Temp - New temp program for Intel Core processors

  1. #2726
    Registered User
    Join Date
    Jan 2008
    Posts
    61
    after days of trying to get an answer, it just doesn't seem like its going to be possible...

    here is the history:

    initial email:

    Is there a chart out there anywhere that shows the TJmax settings for core 2 duo e8600's? I am using core temp 2.7 to monitor my temperatures and there is a variable setting called TJmax that I can change. I can't get an answer from anyone that is consistent. Some people say 90, 95, 100 etc etc. I have searched the intel site but I can't find any info there either. Could you please tell me where i should have this set?

    I'm also curious about the TJmax setting for the e6600 that is in my wife's computer.

    response:

    Since there are many variables to determine the Thermal Junction, we do not have a fixed value. Furthermore, you can review the information in the following link for instructions on how to determine the value:

    http://www.intel.com/support/process.../CS-011039.htm

    Please do not hesitate to contact us again if you need further assistance.

    Sincerely,

    Cindy J.
    Intel(R) Technical Support

    my reply:

    ok, well can you give me some sort of value??? is 85 too low??? is 110 too high???

    their reply:

    We would not be able to give a value range since that depends on many variables. We suggest reviewing the information in the website provided before.

    Please do not hesitate to contact us again if you need further assistance.

    Sincerely,

    Cindy J.
    Intel Technical Support

    my reply:

    well can you help me narrow down the setting??? what variables are involved??? i looked through the documentation you provided and i don't understand 90% of it... im not an engineer and most of that stuff is way over my head...


    their reply:

    The calculation of the thermal junction is not a simple task, and it involves all the variables mentioned in the website. Therefore, what we can do is point you to the formulas and instructions that are properly documented in our website.

    Sincerely,

    Cindy J.
    Intel Technical Support


    my reply:

    so what is the most popular setting that people use for the e8600??? can you just give me a value that i can go by?? are you afraid of your legal department or something?

    by the way, you have to have a degree in calculus to read the documentation provided.

    their reply:

    Perhaps you can check with the software developer if the thermal specification of the processor would be usefule for the utility. For the processor in question, it is of 72.4°C. This is not the value for the thermal junction, but the highest temperature at which the processor could operate without any problems. This is all the information we can provide.

    Regards,

    Cindy J.
    Intel Technical Support

    my reply:

    here is the thing
    maybe you can shed some light on this and help me decide what the appropriate setting is:

    right now, my overall CPU temp is at 41c. this temperature remains the same no matter what i change the tjmax setting to.

    here is where it gets tricky. the following temps are what appears for the individual cores when i change the TJmax setting.

    tjmax 90:
    core 1 = 26
    core 2 = 26

    tjmax 95:
    core 1 = 31
    core 2 = 31

    tjmax 100:
    core 1 = 36
    core 2 = 36

    tjmax 105:
    core 1 = 41
    core 2 = 41

    tjmax 110:
    core 1 = 46
    core 2 = 46

    it makes the most sense that in my case that the tjmax setting should be 100-110 seeing that those temps are the closest to overall cpu temp reading.

    their reply:

    The core temperature will, in most of teh cases, be higher than the overall processor temperature. This is because the cooling solution does not act on the cores directly, but on the heatspreader of the processor.

    The first recommendation would be for you to contact the software developer and verify the software is designed to give accurate readings on processors with two cores. Most of the applications used for this purpose cannot differentiate the readings from the thermal diode and the electronic thermometer, which are the 2 sensors on the processor. Therefore, the readings will be inaccurate and misleading.

    To conclude, since we do not recommend any specific value for the thermal junction, we do not agree with your assumption.

    Please do not hesitate to contact us again if you need further assistance.

    Sincerely,

    Cindy J.
    Intel Technical Support
    (currently clocking)
    asus r3e / 980x / mush ridgies / 2x gtx580 / cs ax1200 / mm pinnacle 24
    loop 1: mcp655 / ek320 / st xt / kl mb-asr3e / bp multi250
    loop 2: mcp655 / st qp320 / 2x st fc580 / bp multi250

  2. #2727
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042
    I think TJMax=100C is a good default value to use and Intel agrees. That is their TJ Target for the E8600. If you use that and your sensors aren't stuck, you'll get some pretty accurate temperatures. Do a CPU Cool Down Test to make sure your sensors don't stop moving at lower temperatures.

    Temperatures aren't that important anyhow. These things run great so enjoy your CPU. That's a lot more fun than trying to get any information out of an Intel rep. Do you think Cindy is human or a bot?

    Edit: Your CPU temp comes from a different sensor and there's no guarantee that it is accurate either. It's just another number unless its accuracy can be proved. The core temp sensors aren't very accurate at lower temperatures so don't base your decision about TJMax on what your idle temps say.
    Last edited by unclewebb; 12-03-2008 at 05:52 PM.

  3. #2728
    Xtreme Cruncher
    Join Date
    Sep 2008
    Location
    Kansas City, Missouri
    Posts
    2,122
    Love this program. Just downloaded the new 2.89.6 and it has many features I love and glad you added to it. Also like the sensor test.
    ~ Little Slice of Heaven ~
    Lian Li PC-A05NB w/ Gentle Typhoons
    Core i7 860 @ 3gHz w/ Thor's Hammer
    eVGA P55 SLI
    8GB RAM
    Gigabyte 7970
    Corsair HX850
    OZC Vertex3 SSD 240GB

    ~~~~~~~~~~~~

  4. #2729
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042
    Thanks HuffPCair. I'm planning to add the actual CPU Load % to the Cool Down Test tonight. It will make it a lot easier to see when people have too much junk running in the background on their system to do an accurate cool down test.

  5. #2730
    Xtreme Addict
    Join Date
    Sep 2008
    Location
    Downunder
    Posts
    1,313
    Way to go hounding Intel there benniebeeker You are contacting the wrong people if you want any useful information though. Tech support are meant to answer as vaguely as possible until you give up.

  6. #2731
    Xtreme Member
    Join Date
    Jul 2008
    Posts
    103
    Love the new 2.89.6! I've always wanted a feature for CPU load. I think progress bars would be unnecessary as I prefer to have as little overhead as possible in my temp monitoring software. Instead of the little rectangular button that lets you cycle between different info given, I'd prefer a settings area where I could tick off which fields I want displayed in the top two lines. So for example, I would like to see on my system:

    Code:
    Intel Quad Q6600         400 x 8.0 MHz (3200 MHz)
    VID = 1.3000             T = 0:06:34

  7. #2732
    Xtreme Enthusiast
    Join Date
    Jun 2007
    Location
    Athens
    Posts
    558
    what about a gadget for the sidebar n Vista ?
    i'am asking too much ? ok :p
    CPU : Q9550 / Board : Asus P5E64 WS Evolution / Ram : 2x1 OCZ D9GTR DDR3 / Vga : HD 4870 / PSU : PPC&C 750W / SSD Ocz Vertex 30Gb / All under Water & Tec's
    Overclockers Wannabe Athens Dept...

  8. #2733
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042
    I've been thinking about expanding the upper section to two lines.
    Good idea about letting a user decide what he wants to see. Maybe a right mouse click and a small pop-up menu where you can select what you want to see for each box. Customizable software is always a good thing. I'll keep that in mind.

    I'm just not a big Vista gadget fan. I had that feature on for the first week but lost interest after that and turned it off. I have a clock on my wall and a watch and I can see the time in the system tray so I found I really didn't need to see one on my desktop. I spend too much time on the computer anyhow. I don't need to be constantly reminded. I also look outside when I want to know the weather. It changes pretty quick around here.

    I know there are some gadget fans out there so I'll continue to think about that too.
    Last edited by unclewebb; 12-03-2008 at 08:26 PM.

  9. #2734
    Xtreme Member
    Join Date
    Jul 2008
    Posts
    103
    Hey man glad I could offer some feedback! Would it be possible to also add a VCore reading? That way I wouldn't need to have CPU-Z running either just for that one value.

  10. #2735
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042

    RealTemp 2.89.7

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

    There was a bug or two in 2.89.6 that if two instances of it were running, they would fight each other and screw up the Load % reported so here's a quick fix.

    I also added more accurate Load Percentage reporting to the Sensor Test. The previous values were Intel theoretical values. I like the actual values better. I think they change depending on what multiplier your CPU has. It's also a lot easier to see if a users computer is truly idle. Some users complain about high idle temperatures when they actually have too much junk running in the background.



    When I ask, "What are your idle temps?", 0.0% load is what you should be shooting for.

    I added a digit after the decimal point because it seemed to be giving me useful information. Even something simple like TaskManager bumps it up 0.5% and changing the update rate of TaskManager bumps it up another 0.5% or so. RT is low stress so it might help isolate and find junk running on your system that you didn't know about. Vista isn't too big of a CPU hog at idle as long as you have it set up properly.

    This version should work better picking up the lower multis when Core i7 is idle. Lots of changes today so let me know if I screwed anything up.

  11. #2736
    Registered User
    Join Date
    Jan 2008
    Posts
    61
    Quote Originally Posted by unclewebb View Post

    Temperatures aren't that important anyhow. These things run great so enjoy your CPU. That's a lot more fun than trying to get any information out of an Intel rep. Do you think Cindy is human or a bot?
    read closely... cindy made typos...
    (currently clocking)
    asus r3e / 980x / mush ridgies / 2x gtx580 / cs ax1200 / mm pinnacle 24
    loop 1: mcp655 / ek320 / st xt / kl mb-asr3e / bp multi250
    loop 2: mcp655 / st qp320 / 2x st fc580 / bp multi250

  12. #2737
    Registered User
    Join Date
    Sep 2008
    Posts
    22
    Quote Originally Posted by unclewebb View Post
    Now the question. Is this important information?
    Should I update RealTemp and make it look like this.
    YEEEEEEES!

    progress bars would be perfect, but numbers is a good upgrade, too!

    Quote Originally Posted by wongnog View Post
    Instead of the little rectangular button that lets you cycle between different info given, I'd prefer a settings area where I could tick off which fields I want displayed in the top two lines.
    totally agree!

    an option to disable the lower part of the realtemp where it says "thermal status" would be useful too.

    I would also like a changelog every time you update realtemp in order to know exactly what changes have been made.

  13. #2738
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042

    RealTemp 2.90 RC1



    It's been a long, long time since the last official release so I want to get something out the door in the next day or two if there are no major problems.

    Here is version 2.90 release candidate 1 (RC1). I like Microsoft speak.
    http://www.fileden.com/files/2008/3/...alTempBeta.zip

    RejZoR and fellow haters of my unlabeled toggle button will be happy with its retirement. I added an extra line of information at the top so we can see what needs to be seen without too much extra work. The APIC ID never changes so it's written into the INI file if you are interested in finding out the order of your cores. Real time VID has been sacrificed but the Min and Max VID are still available in the Settings window for Core 2 based CPUs.

    sakis_the_fraud: I'm holding off on the progress bars for CPU load. When a CPU isn't fully loaded, it's pretty much random data as the load moves around from core to core. I like the single CPU Load number better.

    I would also like a change log every time you update realtemp in order to know exactly what changes have been made
    I'm what you'd call a manic programmer. When I start madly programming, I can barely remember what the hell I've done. It's always a surprise the next morning when I wake up.

    With the official version more or less done, I can now go back through 4+ months of XS posts and try to sort out what all of the new additions are. I love to program but documenting everything is kind of boring.

    The Thermal Status is going to stay as is. It doesn't take up too much room, it would be a hassle hiding it and it should be easy to see so if your heatsink ever falls off, you might notice. The Core i7 runs so hot with the OEM cooler that Thermal Status is becoming a more important feature.

  14. #2739
    Xtreme Enthusiast
    Join Date
    Jun 2007
    Location
    Athens
    Posts
    558
    what about removing this 0x6fb or the timer that i have no idea what they do and put in it's place a vcore reading something like Cpu-Z ?
    also it would be nice to have the ability to hide some line of information with something like this
    Last edited by Vatos_locos; 12-04-2008 at 02:06 PM.
    CPU : Q9550 / Board : Asus P5E64 WS Evolution / Ram : 2x1 OCZ D9GTR DDR3 / Vga : HD 4870 / PSU : PPC&C 750W / SSD Ocz Vertex 30Gb / All under Water & Tec's
    Overclockers Wannabe Athens Dept...

  15. #2740
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042
    0x6FB is the CPUID. If you are looking for your CPU specs at the Intel site it might come in handy. I wish I had data for the 101,000 different motherboards out there that use 1001 different monitoring chips. CPU core voltage would be a big plus. Show me how to do it and I'll get to work on it.

    CPU-Z gets the core voltage more or less correct on the vast majority of motherboards so I doubt RT will ever get that feature.

  16. #2741
    Xtreme Enthusiast
    Join Date
    Jun 2007
    Location
    Athens
    Posts
    558
    oh ok then :p i have no idea of programming just requesting i'am getting greedy
    CPU : Q9550 / Board : Asus P5E64 WS Evolution / Ram : 2x1 OCZ D9GTR DDR3 / Vga : HD 4870 / PSU : PPC&C 750W / SSD Ocz Vertex 30Gb / All under Water & Tec's
    Overclockers Wannabe Athens Dept...

  17. #2742
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042
    If it was simple to do, I would have done it yesterday. CPU-Z has been around for a lot of years and has better connections than I do to new hardware.

    A timer is a handy feature. Now I can see how much time I spent on the computer today. You must not have kids that always want their fair share of computer time.

    It's also handy when you want to keep track of how long you've been doing a stress test for. I put the CPUID label back on so it's a little less cryptic. Thanks for the feedback.
    Last edited by unclewebb; 12-04-2008 at 02:27 PM.

  18. #2743
    Xtreme Enthusiast
    Join Date
    Jun 2007
    Location
    Athens
    Posts
    558
    ok i give up it's nice this way i still love it
    CPU : Q9550 / Board : Asus P5E64 WS Evolution / Ram : 2x1 OCZ D9GTR DDR3 / Vga : HD 4870 / PSU : PPC&C 750W / SSD Ocz Vertex 30Gb / All under Water & Tec's
    Overclockers Wannabe Athens Dept...

  19. #2744
    Xtreme Member
    Join Date
    Jul 2008
    Posts
    103
    2.90 RC-1 looks great! Especially love the top 2 lines. One change I'd like in the System Tray Settings is a choice for Display Core which is Highest Temperature Core. For my system it is always Core 0 or 1, but just in case it ever changes it would be nice if RT would always show the hottest core. It's a setting I miss from that "other" temp monitoring program...

  20. #2745
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042
    Quote Originally Posted by wongnog View Post
    It's a setting I miss from that "other" temp monitoring program...
    Hmmm, in that case, I'll see what I can do.

  21. #2746
    Registered User
    Join Date
    Jan 2008
    Posts
    61
    Quote Originally Posted by unclewebb View Post
    I think TJMax=100C is a good default value to use and Intel agrees.
    i don't think it's 100... if you set the tjmax to 100, the actual core temps are lower then the overall cpu temp which cindy pointed out is the contact on the actual ihs... i think it should be 105 or 110... it makes sense because with those settings the individual core temps are slightly higher than that of the IHS...

    just my 2 cents... i would love to hear your opinion on this theory webb..

    ::EDIT::

    also, after an hour or so of running prime95 with the tjmax at 100, the cores and the cpu temp even out across the board... at least from my experience...
    (currently clocking)
    asus r3e / 980x / mush ridgies / 2x gtx580 / cs ax1200 / mm pinnacle 24
    loop 1: mcp655 / ek320 / st xt / kl mb-asr3e / bp multi250
    loop 2: mcp655 / st qp320 / 2x st fc580 / bp multi250

  22. #2747
    Registered User
    Join Date
    Jan 2008
    Posts
    61
    cool down test @ 4ghz
    (sorry for the unintentional double post)
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	rt.JPG 
Views:	2634 
Size:	47.2 KB 
ID:	90179  
    (currently clocking)
    asus r3e / 980x / mush ridgies / 2x gtx580 / cs ax1200 / mm pinnacle 24
    loop 1: mcp655 / ek320 / st xt / kl mb-asr3e / bp multi250
    loop 2: mcp655 / st qp320 / 2x st fc580 / bp multi250

  23. #2748
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042
    Quote Originally Posted by benniebeeker View Post
    i don't think it's 100... if you set the tjmax to 100, the actual core temps are lower then the overall cpu temp which cindy pointed out is the contact on the actual ihs... i think it should be 105 or 110... it makes sense because with those settings the individual core temps are slightly higher than that of the IHS...
    That's true in theory but you're assuming that your CPU temperature sensor is 100% accurate. It's not. None of these sensors are 100% accurate across their entire operating range and none of them are properly calibrated.

    All sensors have issues. They all have slope error where they change at different rates than the core temperature changes at. None of these sensors are "space shuttle" or military spec. Some are more more like consumer / home hobbyist spec. Many of these are about the equivalent of what you'd find in a $10 science kit.

    Your sensors are not sticking down to 34C. That's better than most.

    Try calibrating them using rge's method outlined here:

    http://www.xtremesystems.org/forums/...postcount=2429

    also, after an hour or so of running prime95 with the tjmax at 100, the cores and the cpu temp even out across the board... at least from my experience...
    That's an example of slope error with different sensors moving at different rates as the CPU changes temperature.

    Here's a good example of how my CPU temp compares to my core temperature readings when wide open.



    At lower temperatures my CPU sensor equals or goes under the core sensors. Depends on the mood it's in. I don't trust any of these things without checking their calibration. The numbers rge has come up with will get you very accurate reported temperatures
    Last edited by unclewebb; 12-04-2008 at 08:10 PM.

  24. #2749
    Registered User
    Join Date
    Jan 2005
    Location
    Las Vegas, Nevada
    Posts
    33
    Still even with 90C on my Q6600 G0 and 100C on my E8200's, they still run much warmer than ambient with a good HSF. The temperatures seem exactly as I would expect them to be. Except possibly the Q6600 G0 spec may be low, could be around 95C or so instead of 90C.
    1. Intel 2500k @ 4.4 / ASUS P8P67 Deluxe / G.Skill 1600 2x4GB / eVGA GTX570 / Hitachi 2x1TB RAID 0
    2. Intel Q9400 @ 3.2 / ASUS Maximus II Formula / G.Skill 1066 2x2GB / eVGA GTX460 SE @ 750/1820 / Hitachi 2x160GB RAID 0
    3. Intel Q6600 @ 3.0 / ASUS P5K-E / G.Skill 1066 2x2GB / Diamond HD4850 / Hitachi 4x320GB RAID 0
    4. Intel E8200 @ 3.2 / ASUS P5Q Deluxe / OCZ 800 Platinum 2x1GB / eVGA GTX260 SC @ 648/2180
    5. Intel E8200 @ 3.2 / DFI P35-T2RL / G.Skill 800 HZ 2x1GB / eVGA 8800GT @ 675/1950
    6. AMD X4 635 @ 2.9 / MSI 785GT-E63 / G.Skill 800 HZ 2x1GB / ATI HD4200
    7. Intel i3 @ 2.13 / Intel PM55 / Samsung 1066 2x2GB / Intel HD Graphics
    8. Intel 2.8E @ 2.8 / ASUS P4C800 Deluxe / HardCoreCooling 2x512mb 4000 / ATI X850XT
    9. Intel P3 800mhz / Intel i815 / Crucial 1x256mb / ATI 9700 Pro

  25. #2750
    Xtreme Addict
    Join Date
    Dec 2006
    Location
    Cochrane, Canada
    Posts
    2,042
    RealTemp uses 100C for the Q6600 G0 based on IR thermometer testing. It's reasonable to assume that the core temperature is ~5C higher than the IHS temperature. The heat spreader does a really good job at spreading heat. I think rge fried an E7200 learning that lesson when he removed the IHS and tried to do some IR testing.

    I have a theory that Intel might deliberately use a slightly different TJMax in the two Dual Cores within a Quad. Screen shot after screen shot I've seen has core0/core1 reporting a slightly higher temperature compared to core2/core3. The difference as the temperature approaches TJMax is usually 0C to 5C higher on the first two cores. It goes way beyond just a coincidence but is difficult to clearly prove. RealTemp is still one of the few programs that orders the displayed data from Quads into its correct physical order. Go check out as many Prime full load Quad screen shots as you can find. You'll be surprised how many are like this.

Page 110 of 180 FirstFirst ... 1060100107108109110111112113120160 ... LastLast

Tags for this Thread

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •