don't think i've posted here yet so here goes.
idle:
http://img80.imageshack.us/img80/417...empidlevv5.jpg
load:
http://img80.imageshack.us/img80/604...emploadgu4.jpg
this is with the top popped. taking it off dropped my temps approx 3c on each core load. idle remained unchanged. i was really hoping that my core temps would be the same after popping the top but oh well. btw, core 0 fails first at 2.9ghz with same 1.4v set as at 2.8 even though it's clearly the cooler core. go figure.
Yah same here, after I took off the IHS it dropped core0 temp about 15C! as it used to be way higher than core1 and now it's even a few degrees cooler than core1 (52 - 54C loaded, vs 55 - 57C core1 @ 1.45v) but still fails sooner than core1.Quote:
Originally Posted by cujo
CPU Information (CPU #1):
Standard CPUID: Family: F, Model: 3, Stepping: 2
Extended CPUID: Family: F, Model: 23, Stepping: 2
CPUID String: 20F32
Processor APIC: 0
Processor: Dual Core AMD Opteron 165 (Toledo)
CPU Speed: 3033 (337 x 9.0)
Revision: JH-E6
Platform: Socket 939
Startup VCore: 1.350v
Maximum Case Temperature = 59C
TDP: 85.0 Watts
.:a BEAST:.
no wonder it does 2.9 at stock voltage :)
latest screenshot 336x9 1.57v
http://weboperative.com/gallery/qdig...per-166ram.png
Looks nice i think i ll try it.
here is screen shot of core temp on my freaking manchester when running prime...is my cpu like on fire or something?
Thats really hot, Im running a Zalman 7000ALCU on this thing.
Its running at 1.52v accoring to bios/cpuz and speedfan.
Speedfan's normal CPU reading(not core one) is reading only about 55C
No its not...Quote:
Originally Posted by GenTarkin
Temp Core its just a non acurate program.
Unless you installed in a wrong way your heatsink there should be no problem since many AMD CPUs have a trothle temps between 65-70C.
No one knows the accuracy of any of these programs. But it's impossible for Core Temp as a program to be inaccurate. All it does is display the value of the thermal sensors. Any inaccuracy comes from the sensors themselves inside the processor.
I thought the throttle of 65-70C was for the TCasemax value?Quote:
Originally Posted by Kain XS
Hi guys
I'm just wondering if you people trust the reading taken from SmartGuardian.
The reason I asked is because Smartguardian is telling me my CPU is running at 72C, yet when I run Core Temp and Speedfan, they both tell me my CPU is running at a much more respectful 35C
How is it possible that Smartguardian's reading is so much higher?
cause its reading the wrong sensor or something. quite ironic for the mobo's own program to read the temp wrong, but smart guardian isnt known as the most trustworthy program by far. if it was really that high, you would not be running stable. same for the previous guy, temps that high are right at the stable threshod, if not higher, so the program is most likely wrong. temps like that wont really do anything more than make the cpu unstable anyways. 80+ is where it is probably a little more worrisome cause of the mem controller more than the chip.
Thanks for that, the PC is running rock solid at readings of 72C. I have been running Prime95 for the last 26 hours without errors.
I guess I should just ignore the readings given by Smartguardian :)
even that... the final result is turning into an innacurate reading using a bad sensor with a conversion.Quote:
Originally Posted by Praz
An accurate reading is to hard to manage...
Once again, comparing Core temp to Tcasemax does not tell you anything.Quote:
Originally Posted by Cra2yMang0
Core temp is the running junction temperature and should be compared to Tjunctionmax.
Tcasemax, the "Tcase", is referring to the temperature on some surface on the IC module.
The Core Temp program in the OP is probably reporting an accurate Junction Temperature. But the problem is that you're not given Tjunctionmax by Intel or AMD.Quote:
Originally Posted by Praz
This is the millionth time I'm trying to convey this in different ways.......
Someone please pick up a book on IC thermal management.
Well, I`ve downloaded the newest core temp and the newest speedfan, coretemp alway`s does like 10 degrees above the speedfan temps.
which program is accurate :rolleyes: ?
Speed fan web site says:Quote:
Originally Posted by Pjoeloe
If speedfan supports temp reading from the cpu's internal measurement you will see this in speedfan:Quote:
Originally Posted by SpeedFan
Speed Fan = Core Temp
Update!!!
New version is out!
Core temp now has its own website, no more posting and downloading it from a forum thread.
Read the 1st Post for the links and info
the link to the prg doesn't work ever for me, i am located in Egypt and don't know what si wrong
can you please host it on filefront or here on the forum for example so i can get it
sorry for bothering but i have been facing this problem since the start of your thread with all versions
Thanks for the great work The Coolest :toast:Quote:
Originally Posted by The Coolest
Your site is now added to my favorites :clap: :banana:
Thanks for notifying me, I'll see if I can get an image up on some other website.
Mirrors are up on the main page of the website
Thak you very much
still can't figure otu why i can't load the original page
My article which discusses motherboard temperature sensors and DTS has been posted just now here:
http://www.overclockers.com/articles1378
I updated the first post with the article link and a "Digg it" link so to increase Core Temp popularity. so please click on it and digg it:
http://digg.com/hardware/Dual_Core_C...itor_EACH_core
I recently purchased a Swiftech Apogee swapping from a DD TDX block as I unfortunatly brought the BEST THING SINCE SLICED BREAD info on the Swiftech web site.
Ive been using this block now for a couple of months, however a couple of days ago I was stress testing my CPU (AMD X2 3800+) using Stress Prime orthos edition, opened up core temp, I noticed that one core temp was around 55oC and the other was upto 72oC , any how this CPU has recently died and gone to heaven, (Mrs's nagging fault) , so I nicked the CPU out of her machine an AMD X2 4400+ , started overclocking it, while stress testing once again with orthos edition of Stress Prime, and Coretemp runing, I noticed I have the same problem with this cpu, one core hanging around 55oC and the other upto 70+oC
I find it very strange that 2 different CPU's can be suffering exactly the same problem.
Any help on this issue would be greatly appreciated.
i cant get a tcasemax value on my am2 3800 X2 EE(65w)
That is quite odd. In idle do both core temps drop? Also what does your motherboard sensor report about CPU temp?Quote:
Originally Posted by Seanie's Show
Maybe an uneven mount?
These values are only readable from the CPU in Rev E chips. I will be adding the max TCM/TDP values for other chips in the next release.Quote:
Originally Posted by WeStSiDePLaYa
Quote:
Originally Posted by The Coolest
ah, ok, nice little prog.
[QUOTE=The Coolest]That is quite odd. In idle do both core temps drop? Also what does your motherboard sensor report about CPU temp?
Maybe an uneven mount?
CoreTemp reports 35-40oC idle temps on both cores, im using MBM5 to monitor temps from the mobo sensor which shows 27oC idle and 41-45oC when priming, unfortunatly although these temps also sound wrong from MBM5, they are more believable than core temp under the water cooling as I find it very difficult to believe one core is reaching 70+oC under water, or even 60oC :stick:
Version 0.94 - 16th November, 2006
- Fix: Temperature log file not created in EXE folder.
- Fix: Temperature log file left "Processor" field blank.
- Add: An actual icon for the EXE/System tray.
- Add: An option to display temperature in Fahrenheit.
- Add: An option to start Core Temp minimized.
- Change: Only list physical processors in CPU selection.
- Change: Improve Intel CPU detection.
- Remove: GCPUID button.
Older versions
See what temps A64Info reports.Quote:
Originally Posted by Seanie's Show
I´m sorry but the file that i downloaded is still the 0.93 version :mad: :mad:Quote:
Originally Posted by The Coolest
Clear your caches.
@seanie: Well, it's the block. I played around with some different blocks and the results were quite interesting, as not all blocks can get you good temps on both cores, or equal temps on both cores...
Maybe you should try to mount the block again mildly and to re-adjust it under Windows were you can watch the temps as you screw down the springs.
It works like a sharm as usual :D
Now a usefull addition would be to display the temp in the tray icon.
That would be really handy. In case of 2 or more cpus you could display the hottest one.
Just a suggestion :)
Edit: Just noticed a little bug. My Sempron 2800+ S754 is being displayed as Venice, but its a Palermo. And it says 69C tcasemax, but I think I remember you saying those Semprons dont have a tcasemax at all.
Dont get me wrong I love this program a lot but it just behaves weird on dualcores, no doubt about it.
After installing the new version (0.94) it was suddenly core #1 that was 3-5 degrees hotter under full load, while I previously had core #0 at a constant 3-5 degrees more. You dont expect me to believe that installing a different version tilted my waterblock or something now do you? ;)
edit: oooh 666 posts! :)
I programmed Core Temp to report the actual core used in the chip, not the CPU's code name.Quote:
Originally Posted by Fr3ak
You have a Venice with half of L2 disabled.
If you just run the 0.94v after using the 0.9.0.91v for a while, that could be the reason.Quote:
Originally Posted by v0dka
Read this about the changes made to 0.93v.
What exactly do you mean, that I am confused by the numbering because in older versions it was 1 & 2 instead of 0 & 1? Isnt the first core above the other one regardless of the numbering?
I think at one time there was a bug, where the temp reported for lets say core A , was acutally core B's temp and vise versa. If you just upgraded from one that was a few revisions old... that may be why you are noticing a change. Unless of course that was a lot of revisions ago.. i didn't check back :p:Quote:
Originally Posted by v0dka
You're right.Quote:
Originally Posted by Diverge
In 0.93v I changed the order of the reported temps as it was wrong.
I also changed the numbering from #1, #2, #3, etc to #0, #1, #2, etc.
So when you update from 0.9.0.91v (changed the long version number too :p:) you notice this difference.
That's what the "CHANGE LOG" is for.
any chance revF's and revG's Tcasemax will ever be displayed?
Maybe sometime.
The problem is that with RevD/E TCaseMax could be read from the CPU itself.
In older and newer CPUs there is no such feature.
it seems strange to me that it would just be taken out though :confused:
perhaps the revF doesn't have it, and revG does... note the max temp:
http://www.amdcompare.com/us-en/desk...=ADO4400IAA5DD
Already checked that filed.
There is a field that is supposed to have some info that should allow you to find the TCM, but for whatever reason it has always been 0 since first RevF chip, same thing with RevG :(
:(
oh well
Should I be worried
my TCaseMax is 59c with a TDP of 85Watts
Im running at 320x9 w/1.45 volts on my 165 opty
Underfully load with orthos my cores are reading Core #0-51c Core#1-59c but mbm5 reads 48c for my cpu.
Is my processor hitting its max Tcase or do I still have +10c to go?
Hi
Am not the great overclocker, as you can se :D
Have i reach the maximum for my cooling.
bios says 66C and core temp says 50C.
Is there someone there ha a link to a guide about memory adjustment.
http://i10.tinypic.com/2cpxlyx.jpg
Your temps are OK.Quote:
Originally Posted by Stoner@Large
I get the same temps as with Everest Ultimate.Is this program blended in Everest?I mean the CPU Diode Read from Everest.Nice program I use it cause it's simple as for the rest I use Everest.Here are my results.They seem pretty accurate.I'm using the stock AMD cooler with lots of dust cause we painted the house so it settled on my caseless PC.
I got a Asus P5B deluxe and a conroe/allendale 4300
From bios and asus probe I get real high temperatures, like 55-60° even idle ... and asusprobe sometimes make "jumps" up to 10° degrees ...
The first time I put all togheter I used the stock heatsink and I thought it was the problem, directly or a bad transfer of heat from cpu to hs .. so today I decided to buy zalman "clip" that make possible to mount my Zalman 7000B on socket 775 and see if temperatures could be better ... but nothing changed
I still get really high temperatures in asusprobe .. but much lower in CoreTemp ... like 20°C of delta or more
What temperature I have to trust ? (I believe the coretemp temperature ... z7000b is correctly installed and it's cold to touch .. I cant get how can cpu be at 60°C)
Do you know if some or all p5b report weird/uncorrect temperatures ? are there bios versions that correct it ? the real temperature isn't the coretemp's one ?
Help me if you can, thanks
and sorry for my bad english
argh, I found this thread searching coretemp on google and didn't see it's on Amd related forum, sorry
Is anyone else finding coretemp to be reading the new e4300 cores too low ???
See HERE for my early findings, they do not correspond to Speedfan
How do I get Coretemp to work in Windows Vista 32-Bit?
This software is reporting rather LOW temperatures for my e4300. PC probe II and everest ultimate edition shows 52 celcius when I'm running orthos blend test , at this point core temp reads 41 on both cores Q.o
If only those temperature readings from core temp were real :P
The cpu is running 3ghz @ 1,35v
I have written a small guide about how to get Core Temp to run on Windows Vista, if anyone is wondering:
http://www.thecoolest.zerobrains.com...opic.php?t=211
This is an odd problem with most/all E4300 chips, and I'm looking into it, but I fail to see how this post has anything to do with AMD CPUs?Quote:
Originally Posted by praesto
http://www.xtremesystems.org/forums/...d.php?t=103638
All that I did was run it in XP compatibility. Simple really.
Which Vista? 32bit or 64bit?
Ahhh, I did it in vista 32 bit.
Looked at your guide now and all I did was step 2. Funny I looked on your site before and didn't see the guide (oh, it's dated today, thats why)
Figured it out myself. It's nice to see you here. :)
Oh, hold on a minute. I didn't have to chose run as administrator. I just chose compatibility mode for xp and it worked fine. Maybe try that with 64bit?
Quote:
Originally Posted by The Coolest
Oh blast, typical me, I wanted to quote that guy whom talked about low E4300 temperatures. Thanks for the link, I'll look there instead of here from now on :D
Version 0.95 - 17th March, 2007
- Fix: Revision F Turion/Turion X2 detected wrong.
- Fix: Core Temp fully loads one core on a dual core AMD CPU.
- Fix: Core Temp crashes on Dual Core Opteron 200 series systems.
- Fix: Core Temp misdetected certain Dual Core Opteron CPUs.
- Fix: Must set Core Temp to run as Administrator in Vista.
- Fix: Can't see Core Temp GUI if Core Temp was last closed minimized.
- Fix: C2D L2 rev. report very low temperatures. (Tjmax to 100C)
- Fix: Core Temp reported wrong temps after a while. (Driver conflict)
- Fix: Core Temp reported Max VID on Intel CPUs. (Now shows current)
- Add: Show temperature in System tray.
- Add: Manual refresh option by pressing F5 or "File" Menu.
- Add: Show "Delta to Tjunction" temperature for Core 2 CPUs.
- Add: Support for latest Intel and AMD CPUs.
- Add: Minimize/Restore option in system tray icon right click.
- Add: Save last window position.
- Add: Prevent 2 instances of Core Temp from launching.
- Change: Rewritten Intel CPU recognition code.
- Change: Core 2 Duo codename change to Conroe-2M from Allendale.
- Change: Pop up bubble-window with temperatures removed.
- Change: Not Beta status anymore.
Nice Work!
what about btrisbane CPU's I wonder :/
might give it a go shortly, thanks coolest
-ed I now see support for latest AMD + intel as a general note there :)
As always..............Thank You for your excellent work :clap: :toast: :banana:
If anyone is having problems with Core Temp crashing on them, please look at the appropriate thread and post there your problem and system spec. I think this way it will be very effective at getting to the bugs faster.
AMD Thread
Intel Thread
doesnt work on Vista 64.
it wont stop loading error messages until you reboot.
otherwise, all you get is 100's of errors over and over.
Till I get the $$$ to digitally sign the driver for x64 here's a workaround:
http://www.alcpu.com/forums/viewtopic.php?t=211
i'm getting 9/10 °c on my brisbane lol ?
AMD has reported the DTS on Brisbanes as being "broken"
I may have found a solution(!) to the BSOD problem in 0.95.
My test system stopped crashing. I need more computers with this problem to confirm the fix is working.
If you want to test the fix, please contact me via one of my IMs.
Thanks,
Arthur
Arthur,
Appreciate this great tool!
I wonder whether it allows users to define an offset to the following formula (mentioned in your article): AMD - CoreTemp = DTS_Value - 49
The DTS on my Brisbane processor gives out readings that are consistent and wrong (probably 14-17C below the actual). If I can calibrate it, I need a way to tell Core Temp how to display right values.
Version 0.95.4 - 22nd August, 2007
- Fix: Core Temp causes restarts/BSODs on start.
- Fix: Core Temp causes hard shutdowns on some AMD systems.
- Fix: System tray icons are black squares on less than 32bit desktops.
- Fix: Some other minor bugs fixed.
- Add: Support new Intel CPUs.
Thanks! I've been waiting for this one.Quote:
Originally Posted by The Coolest
Can you, without getting toio technical, explain the shutdowns?:D Sounds pretty strange to me, but I've experienced it more than once.
^ Thanks so much The Coolest :clap:
The Coolest,
Thanks. I was wondering about the shutdown while stress testing my son's AMD system.
:up:
I'm hoping this issue should now be gone. I haven't tested it much, but it seemed to work fine on a few systems which had the problem before
Could you tell us what you think caused the hard shutdowns? Was it due to non-volatile access to the DTS?
When I open this new version it shuts down and reebots my PC.
The only version I can use is .94
Excellent little program. Opens quickly and give visable and accurate temps. Thanks for sharing it.:up:
Can you please disable auto reboot and see if you get a BSOD? Also, do make sure you're running 0.95.4Quote:
Originally Posted by Kain XS
Now which version of Windows do you have, and what is the error in the BSOD? This would be helpful.
And thanks everyone for the support :)
Not really. I was using it on a AMD X2 4400+ at work the other day and it was reporting core 0 at 16c and 1 @ 9c. Everest gave a more reasonable result of 40c.Quote:
Originally Posted by PhilDoc
For the record I was using .95.
I was using the beta .954 program, havent tried the .95, but will try it today. For me, I was getting the same temps that I was getting with the other programs I tried.Quote:
Originally Posted by HotGore
I ran the non beta version. The temp were certainly not as off as yours but core 1(2) was 10C lower than pcwizard. After exiting out of Core Temp the reading for core1 in pcwizard dropped 8C. Strange
Thanks :)
Version 0.96 - 31st December, 2007
- Add: System tray icons configuration.
- Add: Support for Intel's latest CPUs.
- Add: ClearType font support.
- Add: Time\Date column to the log file.
- Add: Always on top option.
- Fix: Core Temp does not proprerly display temperature and CPU select fields.
- Fix: Log file layout.
- Fix: Hide\Show does not properly Minimize\Restore window.
- Fix: Core code names aren't properly detected. (Intel)
- Fix: Incorrect\Missing CPU rating when not using stock\maximum multiplier. (Intel)
- Fix: Wrong CPU clock is reported when using EIST\non-maximum multiplier. (Intel)
- Fix: Wrong detection of Merom based CPUs.
- Fix: Single core Athlon 64s and Semprons reading -49C constantly. (Rare)
- Fix: C2D M0 rev. report very low temperatures. (Tjmax to 100C)
- Fix: Layout and display of tool tip text in the system tray. (Now shows a core per line)
- Change: Log the start and end of a session even if logging is disabled.
- Change: Settings and Systray settings windows appear in center of main window.
- Change: Rename "Tjunction" field to "Tj. Max" to make things clearer.
Happy new year to everyone :)
happy new year bro
It's just awesome that you continue your work on the program.
Certainly a "must have" and the "standard" when comparing processor temps :up:
Will we see suport for the K10 Cpu-z sone? happy new year!
my 5000BE on a msi k9a2-cf mobo (790x chipset) reads 5-15c idle and 15-25c load. which is ~20c less than what is read by the bios. which im hoping is accurate, since i really have no programs that really seem to support these chipsets yet. so is it the cpu or the chipset thats causing the false numbers?
Have looked at Phenom docs and it doesn't seem to have the DTS register that exists in all K8s.Quote:
Originally Posted by BadNizze
There is some sort of register with a temperature, but it doesn't report an absolute temp (meaning it doesn't report the actual CPU temp). It is a single value for the whole CPU, and it's useless as it doesn't tell you the actual temperature of the CPU.
The Brisbane chips, for whatever reason always report incorrect temperatures. I don't know why, but it's just the way it is :(Quote:
Originally Posted by Manicdan
Core Temp doesn't do anything with the data it recieves from the processor (which is supposed to be accurate in the first place) and just displays it to the user in an understandable way.
HW Moitor works on my X4, but as you suspected it´s not reading from each core, but from an "average" temp from all cores. Sorry for bad spelling hope you know what i mean!Quote:
Originally Posted by The Coolest
Version 0.96.1 - 15th January, 2008
- Add: Support for *Phenom and "Barcelona" Opteron.
- Add: Support all K8 processors.
- Add: Support Intel Celeron M 500 series.
- Fix: Wrong CPU speed on Intel ES chips.
- Fix: Fix M0 Tjunction max detection.
- Fix: Intel QX6850 CPU's rating shown as QX6650.
- Fix: Intel E6550 CPU's rating shown as E6650.
- Fix: Intel Celeron M 400 series detected incorrectly.
- Fix: AMD Opteron for Socket AM2 reported as Santa Rosa instead of Santa Ana.
*Please read this.
Version 0.97 - 5th March, 2008
- Add: Vista x64 support - All drivers are now digitally signed!
- Add: Logitech G15 keyboard support - see ReadMe!.txt for details.
- Add: Support 45nm LGA775 Xeon series.
- Fix: Opaque background in system tray wasn't 16x16 pixels.
- Fix: Yonah based CPUs incorrectly detected.
- Fix: 1333FSB Dual Core Conroe based Xeon incorrectly detected.
- Fix: Dreaded cycle of "Driver can't load" messages if driver failed to load.
- Fix: Socket AM2 Athlon-FX was not recognized.
- Change: Core Temp layout on a Phenom.
Downloaded the update (since the previous version gave me the exact same temp on all four cores at the exact same time all the time) and this is what I get. If I hover my mouse over the four temps in the taskbar the tooltip says 33C for all four but as you can see only the first one says 33 the rest are 0, and in the coretemp window itself, well... I sure hope my CPU isn't that hot ;) When i first load up coretemp it reads 33C then it jumps to that crazy high number after a second.
Not ripping on the program, have used previous versions and been very accurate and helpful, but this one seems to not agree with my phenom too well.
I included a screenshot of the previous version for reference.
That looks like a nasty bug. I will try to sort it out in the next few days (depending on time availability).Quote:
Originally Posted by SparkyJJO
Thanks for the head's up.
Sure thing, keep up the good work! :up:
i ran it with my 9500 and i had a normal reading for a second and then it jumped to the crazy number, i changed the default format to farenheit, and it stayed at 91, while all the other numbers in the taskbar were still sitting at the celsius reading(correct temp). Hope it helps u out a little, great program , and i'm so happy there is phenom support now.
Thanks for the work The Coolest. :up:
There's a bug in 0.96.1, that it didn't pick up any CnQ changes apart from VID. It did however display the temperature fine. There were random high temp. values thrown back at times though.
Attachment 73582
CnQ:
Attachment 73584
High temp. jump:
Attachment 73583
With 0.97.1, there are many bugs. Firstly, it doesn't display the temp. correct at all, only one tray icon displays the temp. properly on one core.
Secondly, it shows one core temp (I know there's only one register feedback).
Thirdly, it picks up VID/Multiplier change with CnQ/MSR editing, but doesn't change the overall speed, so works out the wrong HT ref. Maybe you're working out the CPU speed using the IMC/RAM speed instead of CPU multi*HT ref.?
Randomly shows this temp:
Attachment 73585
Mostly shows these temps and this HT bug in CnQ mode (or if CPU multi is changed after boot):
Attachment 73586
Why do i get 0.96.1 instead of 0.97 from the links on first post?