[Desertfish]

As some of you might have read, I started having some problems when upgrading both the motherboard and CPU in my system (see this thread for more info: http://www.computer-juice.com/forums...pro-why-18574/). Anyway, after finally managing to get my E8500 CPU and Asus P5K Pro mobo speaking the same lingre, I did a clean install of WinXP-SP2, installed all the drivers and software, only to come to the horrible conclusion that the software I'm planning to use the most (i.e. Corel Painter X) was causing BSOD's when I launched the application...

Now that is great... After forking out £200 for a new mobo and CPU, I'm now not able to use the application for which I paid £285 about 3 monts ago... Brilliant (Duh!). Since the problem was only happening when I launched Painter, I thought it must be some sot of a config issue, possibly the app conflicting with some driver. So I installed XP again and then immediately the Painter app (without installing any device drivers) and guess what... Same BSOD pointing to fault in win32.sys and an error code of 0x000007f... So the Windows kernel got corrupt already... Me thinks not! Especially because Painter was not showing this behaviour on 3 different systems I tested it on (well, I got my own laptop and a few machines to play with at work...). I thought is was clear that there was something wrong with my new hardware, either the motherboard or the CPU. Since there is a lot more chance of the mobo being at fault then the CPU, I suspected Asus had sent me another dodgy product...

Out of interest (please don't ask me why) I decided the test the behaviour of Painter X by installing SP3 and - oh my God, it worked... Until I found out that I could use the application only once... That's to say, whenever I booted my system I could launch Painter and work with it for as long as I wanted. However, when I closed the app and launched it again, my system would hang... The only way to get around that was to reload the old girl... Done some searching around on t'interweb and came across some info about "minidump" files and how to analyze the info using WinDbg... Interesting... One long boring nightshift at work spent on seaching the web for "Debugging for Numpties" info, I decided to have a go at it... The results were interesing... Where the BSOD constantly told me that an issue was caused in win32.sys, the debug session showed that it "ntoskrnl" was being affected by a process that was having access to parts of the Windows kernel that is was not supposed to have access to. Since Windows was not happy about this at all, it issued an instant system kill... WinDbg told me that error 0x000007f with a first parameter of 00000008 meant that there was a hardware problem and nicely pointed it to my CPU... "Go and talk to Intel..."

I found this a very weird development but followed the advice nevertheless (sort off...) I googled "0x000007F 00000008" and the first link brought me to the Microsoft Support site, explaining the meaning of the error code and confirming that without a doublt, this was a CPU problem. The advice given was to test the CPU in a different system and test a different CPU in your current one... Not something that most of us can do, but fortunately, I have many spare parts laying around so I decided to take Microsofts advice (sort off...)

I already knew that my old mobo and old CPU were working fine and that the new mobo and new CPU were causing problems. All I needed to do was to swap the CPU's and test the two combinations that were left... And I was in for a SHOCK...!!! Putting the E8500 processor in both my Asus motherboards produced the same result. As soon as I launched the application (again using WinXP-SP2) I had the same BSOD as before. However, when I put my old CPU in both motherboards (and E2160 running only half the speed of the E8500) not only were all my problems solved, the little bugger of a cheapskate processor was also outperforming its big bro... ! Not that I ran any benchmarks (so far), but me and a friend of mine who popped in halfway the afternoon, clearly noticed that Windows was booting a hell of a lot faster with the E2160 and that resource hogs like MS Office, Adobe Photoshop CS3 and Illustrator CS3, as well as Corel Painter X were loading noticeably faster...

Now it seems I have to go and talk to Intel about this cause the question remains: "is my E8500 dodgy or are there problems with BIOS support for this CPU...?"



Anyway... sorry for the long post guys, but this was intended as an illustration of where troubelshooting can take you, and how (at times) all logic can be lost when a computer is showing unwanted behaviour...

[Mike0001]

Oh boy!

[Bubba]

I'm sad that the only response to this was "oh boy". I would have hoped that someone could shed some light upon this. I sounds to me like a bios problem (but i'm a numpty so take THAT with a grain of salt), as I have had some problems myself. Kind of the same but different all at the same time lol. I myself am at wits end. I'm going to try to describe everything the best I can and see what Carbon has to say about it. Looks like this part of the forum isn't the place to do it lol.

[Desertfish]

Actually Bubba, this turned out to be a very long story... I've only managed to fix this about last week.

What exactly casued the problem, I'm not sure off, but because my E8500 CPU turned out to be working fine in two different motherboards, while it would cause one problem after another on my P5K Pro board (no matter what BIOS revision was loaded) I would say it is a BIOS problem as well.

I've been searching on the internet for people with a similar setup but the closest I got was others having E8500's running in P5K Premium and P5K Deluxe boards. So I did a comparison of all the BIOS revisions that are available for the whole P5K range. It tuns out that where there is a new BIOS release for the Premium and Deluxe boards about every 3 months, Asus is releasing BIOS revisions for the Pro versions about twice as fast.

I can only suspect that Asus is having difficulty with the support of certain processors on this particular motherboard. If that is true though, I don't see them admitting it.

Cheers:
Chris.