New version OCCT stability test

[p2501]

Kai, the zip archive worked for me as well as other ppl, so maybe something is wrong with your dl-manager? I downloaded via FreeDownloadManager..

cantankerous, I believe the OP lists the official downloads.

[C0ncrete]

Cool going to give it a whirl.

Cheers.

[kiwi]

Thanks Tetedeiench for your great work

Is there an official link to the 2.0.0a? I know there was a beta posted above, is this the same one or have you now made it official and not beta?

http://www.ocbase.com/OpenBeta/OCCTPT2.0.0a.b01.zip

[G80]

Im gonna give it a try. Is the tests more cruel then the other programs? how does it operates more quickly?

[Tetedeiench]

There's only one download link to OCCT

http://www.ocbase.com/download.php?fileext=exe for the .exe version
http://www.ocbase.com/download.php?fileext=zip for the zip version

You'll always get the latest stable version (now is 2.0.0a, which is 2.0.0a.b01 promoted to stable).

As for the minute-long DL... seems my provider is messing a bit up at the moment, it should be better soon

[cantankerous]

Ok I had it when it was labelled a beta. I just re-grabbed it to get the final.

I take it this is the one and only version to use now? No more 1.2.0b06/b07? I had them both cause both offered different things like the slider etc. If you say 2.0.0a is the only one to use now I will erase the others off my drive.

Thanks again!

[lior307]

Great News ( i don't remember the oldest OCCT showed the CPU Temp) .

[g1raffe]

Downloading now, I much prefer the OCCT Interface to Orthos, and the ugly Prime95!

[3oh6]

the .ZIP is working great on DFI 790FX board with a Phenom in Vista x64. working perfect for finding fully stable and un-stable setups nice and quick. really appreciate the update

[Richard Dower]

Great stuff!...Tetedeiench, can you update XS members as and when you have any new betas or official releases here?, as and when they are ready etc.

[Tetedeiench]

Great stuff!...Tetedeiench, can you update XS members as and when you have any new betas or official releases here?, as and when they are ready etc.

I'll try to It means more work for me (as i usually release the changelogs of every betas in french) but i'll give it a shot.

Next beta release should be in the next few hours. It'll probably be promoted stable quickly, as it deals with a problem with speedstep (graphs are generated in some cases with the speedstep frequency on it... i'd rather have the "real" frequency on it).

[Spyrus]

It cannot detect the overclock percentage very well.

[Tetedeiench]

It cannot detect the overclock percentage very well.

OCCT will report the current CPU Frequency. If you have Intel Speedstep enabled, then OCCT will display the frequency modified by Speedstep.

Speedstep is an intel feature that changes the CPU Multiplier "on the fly" when your CPU is unused, in order to save energy. Your CPU will be downlocked, thus using less power.

The CPU FSB will remain the same, though.

When a test is launched and the CPU is busy (i.e. after a minute in Auto and Custom (fixed) tests, and immediatly after launching a Custom(Infinite) test), the CPU Multiplier will be restored and OCCT should display the real overclock %.

Does this explain your problem ? If not, just tell me what is your setup (CPU, motherboard) and tell me the values OCCT detects (a screenshot is more than enough).

[Epsilon84]

I agree with the OP, I've always preferred OCCT to Orthos - in 30 minutes it can pick up errors that may take many hours to occur in Orthos.

[Spyrus]

OCCT will report the current CPU Frequency. If you have Intel Speedstep enabled, then OCCT will display the frequency modified by Speedstep.

Speedstep is an intel feature that changes the CPU Multiplier "on the fly" when your CPU is unused, in order to save energy. Your CPU will be downlocked, thus using less power.

The CPU FSB will remain the same, though.

When a test is launched and the CPU is busy (i.e. after a minute in Auto and Custom (fixed) tests, and immediatly after launching a Custom(Infinite) test), the CPU Multiplier will be restored and OCCT should display the real overclock %.

Does this explain your problem ? If not, just tell me what is your setup (CPU, motherboard) and tell me the values OCCT detects (a screenshot is more than enough).

Nope i have disabled the cpu features. Later on this day i will post screenshots

[Praz]

It cannot detect the overclock percentage very well.

Displays correctly on the P35 and X38 DFI boards. With C1E enabled or disabled. And by the way, thanks for the update to the program.

[Tetedeiench]

As promised :

OCCT PT 2.0.0b.b01 OUT !

Changelog :
* It should be much harder to trick OCCT with SetFSB and such programs to get a higher frequency displayed in the graphs
* Added japanese and bambara (malaysian/indonesian language). Italian language was updated as well

Enjoy !

http://www.ocbase.com/download.php?fileext=beta

[cantankerous]

Great thanks! I really appreciate you personally visiting us here at XS with these updates.

Would you say 2.0 should totally replace the use of 1.20b06/b07 in terms of accurate results while testing?

[kiwi]

Tetedeiench, questions for you

What do the error codes mean? I remember code 4 was calculation error but code 1 and other?

Why does OCCT sometimes terminates closing window without showing any error? So it simply exits in the middle of testing

[p2501]

Tetedeiench, questions for you

What do the error codes mean? I remember code 4 was calculation error but code 1 and other?

Why does OCCT sometimes terminates closing window without showing any error? So it simply exits in the middle of testing

I know you didn't ask me but Code 1 simply means that one of the other cores failed, for me it was mostly Code 4 on another core. But I don't know why it doesn't display the error code for the failed core instead of another that was fine. Go to the occt directory and search through the ThreadX.txt and you'll find what you're searching for.

I don't know about the program closing in the middle of a test run. I'm guessing unstable overclock?

[Tetedeiench]

Tetedeiench, questions for you

What do the error codes mean? I remember code 4 was calculation error but code 1 and other?

Why does OCCT sometimes terminates closing window without showing any error? So it simply exits in the middle of testing

I'm working on getting the error codes on the main frame - it should be available soon.

The error codes are explained in the file OCCT.chm shipped with OCCT - this help is just too basic right now for me to include a éhelp" button in OCCT

As for occt "crashing" : yes, i guess it's related to an unstable overclock.

As for the error codes :
1- Thread stopped because another thread has stopped for some reason (mostly an error 4)
2- User interrupt
3- Not enough memory to run the test (should never happens as the GUI is taking care of that)
4- Calculation error detected

And yes, this version should replace 1.2.0.b06...

[Jon Boy Deluxe]

I have used OCCT for ~2 years but have found it to be an invaluable tool.

However, I still use prime too just to be sure everything is rock solid, I have had an occt stable oc fail in prime and vice-versa.

[Richard Dower]

As promised :

OCCT PT 2.0.0b.b01 OUT !

Changelog :
* It should be much harder to trick OCCT with SetFSB and such programs to get a higher frequency displayed in the graphs
* Added japanese and bambara (malaysian/indonesian language). Italian language was updated as well

Enjoy !

http://www.ocbase.com/download.php?fileext=beta

Legend!

[Spyrus]

My mistake it is ok.

[virtualrain]

How does OCCT work?

When I used it, based on the way it loaded my system and the noise coils would make under load, it seemed to cycle through calculations almost exactly the same way Prime95 goes through FFT calculations.

Is it not using the same or very similar calculation methods as Prime95?