"Core Temp" - CPU Temerature monitoring tool

[sRHunt3r]

Can core temp properly report dual Xeon hexcores?

[The Coolest]

It should, otherwise I'd like to know

[sRHunt3r]

It works great ....... good readings all 12 cores displayed and frequency is reading correctly

Thank you!

[The Coolest]

Any last quick requests? New version is going through final testing right now.

[The Coolest]

A small update

I've been playing around with Media Center a little, and came up with Core Temp MC, an add on for Windows Media Center.
Check it out here:
http://www.alcpu.com/CoreTemp/addons.html#CoreTempMC

[The Coolest]

Version 0.99.7.0 - 26th June, 2010

- New: Implemented Windows 7 taskbar button features. (Under the Options menu)
- New: Clock modulation detection for Intel processors.
- New: Proper support the Phenom II X6 and Athlon L100 series.
- New: Support for the E3000 Celeron series.
- New: Displays the processor's manufacturing process. (Lithography)

- Fix: Intel frequency detection.
- Fix: Frequency doesn't update after exiting from Sleep or Hibernation. (Intel)
- Fix: Reading and logging intervals in Settings can't be changed.
- Fix: Minor Intel processor recognition issues.
- Fix: On some systems the System Information window would flood the screen with errors.
- Fix: G15 plugin restarts after resuming from sleep even if it was previously disabled from the menu.
- Fix: Overheating protection on AMD processors bug.

- Update: Rearranged the main GUI and added new information.
- Update: Added missing Intel Socket names.

There is also a new Media Center addon to see your CPU status here:
http://www.alcpu.com/CoreTemp/addons.html

[some_one]

[The Coolest]

@ some_one
Thanks, it's now fixed, you can download again.
0.99.7.1

[some_one]

IMHO reporting SLFM and FSB dynamic switching comes with it's own problems. Whether to report an "effective BCLK with technically correct multi" or a "technically correct BCLK with effective multi".

While using an "effective BCLK" may show a technically correct multi it does have some disadvantages. For instance when showing a multi of 6x the user now has to check BCLK or Freq to see if it's say 133 x 6 or 266 x 6 whereas using an effective multi (3x) it is immediately obvious. Also note that in the above screen shot showing a 5x multi (0.99.7.1 now shows 5.0 too), technically it is a 10x multi with SLFM enabled 100% of the time. You can see how this can get very confusing. At least your not having to show FSB which would make things even harder while dynamic switching is involved.

Though neither method is perfect I would have thought the "effective multi" would be the better way to go, after all this is the ratio that you are calculating using the performance counters. Just a personal thought and whatever you decide do remember the saying "you can't please all the people all of the time".

[The Coolest]

Yes I noticed that the multiplier jumps around, and goes lower than the lowest possible multiplier.
I still think that I'm going to stick to this method, because as a user, you don't sit and stare at the FSB x multiplier the whole time. You simply glance at it once in a while.

[some_one]

Yes I noticed that the multiplier jumps around, and goes lower than the lowest possible multiplier.

Well, if you take it as an effective multiplier then it really is correct. But I hear you, after all the primary thing here is reading core temperatures.

[unclewebb]

IMHO reporting SLFM and FSB dynamic switching comes with it's own problems.

I definitely agree with that. Internally, the multiplier and SLFM mode can be rapidly cycling so it's a compromise no matter what you report and depending on the timing of when two different programs read data from the CPU can result in slightly different results.

I just switched RealTemp too since

6.0 x 133 MHz = 798 MHz

is technically more accurate than

3.0 x 266 MHz = 798 MHz

Unfortunately, the correct way can make it more difficult to see the data with a quick glance. Doing things the correct way means now you have to look at both the bus speed and the multiplier. CPU-Z just reports the 3.0 effective multiplier and reports a constant bus speed so maybe I'll just go back to that to keep things simple.

[The Coolest]

I've talked to Franck and he said that handling SLFM is a pain, too.
I personally am going to stick to the method I have going right now. I think that it doesn't make sense to change the frequency detection or the way the frequency is reported every other day, it gets old
Right now all of us have it working one way or another, and overall, reporting correct clock speeds. That's the most important thing.

[JoeBar]

May this cooperation will result in a new prog, Real CoreTemp maybe...?

[The Coolest]

Been pretty quiet this week.
Haven't seen any more reports about incorrect clock speeds or anything else for that matter
Any requests/suggestions for future versions?
Also are the K series CPUs recognized properly in the new version?

[panosk]

Why does not report the correct frequency?

[IMG] Uploaded with ImageShack.us[/IMG]

[The Coolest]

Is that speed shown constantly or does it vary?
Because your CPU could be entering a low power mode and so the speed is lower.

[panosk]

Is that speed shown constantly or does it vary?
Because your CPU could be entering a low power mode and so the speed is lower.

the speed varies from 13 to 17 and I have disabled all the power saving options in bios.cpu-z shows the correct frequency

[The Coolest]

Just out of interest, could you try the latest RealTemp and see what frequency it reports?

[Ghostleader]

I can report that 0.99.7 works okey with my i7 875k, now the multiplier stays steady when I have set it fixed in the bios with the powersaving disabled, with 0.99.6 the multiplier went up and down during idle and load as if powersaving was enable in bios but it wasn´t.

Great work

[The Coolest]

Great
Is it recognized correctly as well?

[unclewebb]

panosk: It's usually C1E that is enabled that causes your multiplier to hunt up and down at idle. CPU-Z ignores this but that's what is happening within your CPU. Turning something off in the bios does not guarantee that it gets turned off in the CPU. Some settings get ignored.

RealTemp has an option to turn C1E off in the Settings window if you want your multiplier to be steady at idle.
http://www.fileden.com/files/2008/3/...alTempBeta.zip

[exela]

Is working great. Thanks !

[The Coolest]

Version 0.99.7.3 - 5th July, 2010

- New: "Start Core Temp with Windows" feature on Windows Vista and 7.

- Fix: Taking a screenshot would not capture the whole window in Vista/7 with aero mode.
- Fix: Windows 7 taskbar Cycle mode only shows up after the first interval has elapsed.
- Fix: Overheating protection UI bug.
- Fix: The Mini-mode menu was not checked when Mini-mode was activated.
- Fix: Corrected the placement and size of the load label on Phenom based processors.
- Fix: Order of appearance of the system tray icons in Windows 7.

- Update: The appropriate G15 menu is now checked depending on the status of the G15 plug-in.
- Update: Added support for 12 core processors to temperature offset adjustment window.

[kpablo]

When you set it to start with windows... at the start always shows a message "Another instance of Core Temp is already running!"

Core Temp 0.99.7.3 64bit version.