IB owners, read this

[MPiland]

Read this post by HardwareDecoder at this site:

"So I went back to 4.4ghz @ my prime 95 17hour stable settings. For some reason I ended up in event viewer and saw it had hundreds of WHEA event 19. in the details it said problem with processor parity. I had no idea what this meant so I went on intel live chat and asked them about it, they said it was not a good thing and to run the intel processor diagnostic tool. I set my chip back to stock settings 34 multiplier 100 bclk. IT FAILED!!! apparently the chip thinks it has a 101 bclk? I had no idea that this chip was defective (seemed to be running fine) Anyway I'm gonna RMA it. I'm a little annoyed for sure. I never had the volts over 1.290 which is well below damage range so I guess it was defective since I got it. Just glad I found out now instead of later. Oh well I ordered another one and It's already $5 cheaper with $20 gift certificate until tomorrow.... oh and cpu-z shows a 100 bclk... so idk. But if I were anyone with a new IB chip I would run that test to make sure it's not jacked up also check out for WHEA 19 warnings in event viewer. You can just type event viewer in the start bar and open it up go to windows logs > system. you can also filter the results to only show warnings and errors."

And then a couple pages later he tells another member to try running an Intel CPU test. MaFi0s0 did so and he wrote "Intel diagnostic failed, said I was running 3.5Ghz pass, but 99 base clk fail, cpuz and bios sys 100bclk".

Check your chips, folks. They could be bad stock if you're getting bad temps and OCs.

 

[Ferzerp]

That event is logged when you're almost stable, but not quite. The processor is reporting to the OS that it is experiencing errors.

 

[MPiland]

Yeah, I know what the error is, but most don't think to check the event log. You'd think running P95 for 18 hours would mean you're stable, but that might not be the case.

 

[Idontcare]

This is a great tip! I'm going to include it in the OC'ing sticky. The tricky part about stability testing is avoiding the false postives (i.e. avoid thinking your system is stable when it really isn't). Nobody wants silent data corruption.

 

[MPiland]

Exactly Glad to be of service!

 

[2is]

So if I had say, 101 set as my bclk in the bios, it would fail the test?

 

[DigitalWolf]

So if I had say, 101 set as my bclk in the bios, it would fail the test?

It shouldn't really. Parity error should be related to memory or memory controller but 101 shouldn't really be an issue... all by its self. To me that would be pretty common with some of the speeds people are trying to overclock memory to on Z77... but you could likely cause it by clocking the cpu or bclk to high as well.


Personally I always watch my event logs after noticing a few issues that were later fixed by Microsoft. So its definitely a good thing to bring up (checking event log now and then).

 

[RampantAndroid]

Ooh, thanks! Bookmark.

 

[2is]

Well my event log is clean but it did fail the intel diag tool when it came to base clock since I do have it set to 101. I'll set it to 100 and try again.

 

[2is]

So it appears it doesn't matter where I set my bclk, it will fail that test. Even at 100. Funny thing is that regardless of where I set my bclk at (i've tried 100, 100.1 and 101) it still shows base clock detected at 100MHz.

Seems more like an issue with the diag tool than the CPU to me.

 

[MPiland]

Yeah, i have a ticket in with Intel right now about failing the bclk test at stock. They said they don't know of IPDT not being compatible with IB, but they're looking into it and will let me know. BTW, there will be some errors that are ignore-able in the event viewer. I've found quite a bit. Especially with some of the Asus software.

 

[2is]

Yeah, I have a few warnings that have nothing to do with the processor

 

[MPiland]

Same. I'm still working on fixing them anyway. Don't want random BSOD's cause of stupid files.

 

[Avalon]

I'm all clear. Thanks, though!

 

[nyker96]

I'm still unsure what this means. So if you get parity error which is logged in the event viewer, it means you memory is unstable? If so can you detect a failure running memtest86?

also, if you system passed P95 for a long time, does that mean your error is NOT fatal and somehow self corrected via some mechanisms?

finally, how do you know it's a processor error when it could also be say a motherboard or mem error that's causing this?

not trying to force you answer questions you might not know the answer to. Just wondering out loud, what all this might mean? somehow I'm confused how this point to a cpu failure.

 

[tweakboy]

For stability checking. 48 hours straight will guarantee you pretty much if its stable or not...

 

[BallaTheFeared]

48 isn't enough, needs a year and 13 weeks of prime95 or it's unstable imo.

Jokes aside, unless you're testing for max temps stability test in the programs you use.

Checking the event viewer after each session seems like a great idea as well.

 

[guskline]

48 isn't enough, needs a year and 13 weeks of prime95 or it's unstable imo.

Jokes aside, unless you're testing for max temps stability test in the programs you use.

Checking the event viewer after each session seems like a great idea as well.

That's funny!

 

[MPiland]

I'm still unsure what this means. So if you get parity error which is logged in the event viewer, it means you memory is unstable? If so can you detect a failure running memtest86?

also, if you system passed P95 for a long time, does that mean your error is NOT fatal and somehow self corrected via some mechanisms?

finally, how do you know it's a processor error when it could also be say a motherboard or mem error that's causing this?

not trying to force you answer questions you might not know the answer to. Just wondering out loud, what all this might mean? somehow I'm confused how this point to a cpu failure.

Some errors will tell you directly if it's an error with the CPU or not

 

[amenx]

48 isn't enough, needs a year and 13 weeks of prime95 or it's unstable imo..

Yep. And when the test is done chuck it out and buy Haswell which will be out by then.

TBH one failed chip does not mean very much. Once I see 2 or 3 more within next couple weeks, only then will I raise an eyebrow over IB.

 

[BallaTheFeared]

Remember when that EU forum/site said anything above 1.35v would kill SB within hours?

Trolling for hits is an art.

 

[DigDog]

i love how i have no idea what anything in the thread means :?

 

[MPiland]

Yep. And when the test is done chuck it out and buy Haswell which will be out by then.

TBH one failed chip does not mean very much. Once I see 2 or 3 more within next couple weeks, only then will I raise an eyebrow over IB.

I'm waiting on Intel to tell me if their test is faulty or if I have a bad chip. They're looking over the test results. At stock, I failed the bclk test which IMO shouldn't happen. I'll keep y'all updated.

 

[Cerb]

It shouldn't really. Parity error should be related to memory or memory controller but 101 shouldn't really be an issue... all by its self. To me that would be pretty common with some of the speeds people are trying to overclock memory to on Z77... but you could likely cause it by clocking the cpu or bclk to high as well.

What about 101*44? Your CPU has memory inside of it, too. Unlike main RAM, it can be, and is, checked for errors.

I'm still unsure what this means. So if you get parity error which is logged in the event viewer, it means you memory is unstable?

The error should have that information, though it may not always know (unknown hard error = fun!).

Given how new IB is, however, it could be a BIOS or Windows bug, too.

i love how i have no idea what anything in the thread means :?

http://en.wikipedia.org/wiki/Machine_check_architecture
http://msdn.microsoft.com/en-us/windows/hardware/gg463286

 

[Nemesis 1]

Read this post by HardwareDecoder at this site:

"So I went back to 4.4ghz @ my prime 95 17hour stable settings. For some reason I ended up in event viewer and saw it had hundreds of WHEA event 19. in the details it said problem with processor parity. I had no idea what this meant so I went on intel live chat and asked them about it, they said it was not a good thing and to run the intel processor diagnostic tool. I set my chip back to stock settings 34 multiplier 100 bclk. IT FAILED!!! apparently the chip thinks it has a 101 bclk? I had no idea that this chip was defective (seemed to be running fine) Anyway I'm gonna RMA it. I'm a little annoyed for sure. I never had the volts over 1.290 which is well below damage range so I guess it was defective since I got it. Just glad I found out now instead of later. Oh well I ordered another one and It's already $5 cheaper with $20 gift certificate until tomorrow.... oh and cpu-z shows a 100 bclk... so idk. But if I were anyone with a new IB chip I would run that test to make sure it's not jacked up also check out for WHEA 19 warnings in event viewer. You can just type event viewer in the start bar and open it up go to windows logs > system. you can also filter the results to only show warnings and errors."

And then a couple pages later he tells another member to try running an Intel CPU test. MaFi0s0 did so and he wrote "Intel diagnostic failed, said I was running 3.5Ghz pass, but 99 base clk fail, cpuz and bios sys 100bclk".

Check your chips, folks. They could be bad stock if you're getting bad temps and OCs.

Well from experiance . It looks like your M/b is at fault and not your CPU. I will keep up with this thread until you reinstall the new cpu.