____Intel RMA Saga____RESOLVED! (sort of)_____

[Blain]

I've got a CPU that has a problem with its internal cache, causing Windows 98 registry errors.
I've only had this "retail" CPU for a short time. I did not "buy" this processor. I got it in a trade deal here.
Therefore I don't have any receipt showing the original purchase. Does the bar code and information on it count as their proof of purchase?
Or will I need some type of official receipt from the person that I got it from?
I'm going to call their customer support today. I'd just like to have as much information going in as I can get.

BTW Even though I got this CPU in a trade deal, I am the first to use it. It was sealed in the retail box when I got it. I've just been running it at the 100MHz default speed.

Thanks

 

[Gunbuster]

I had no problem geting a new fan for my retail p3 500 when it went bad


Just have the cpu handy or all the numbers off of it (serial, sl, everything).

 

[Blain]

Well getting a new CPU for one with an internal cache that's gone bad may not be as easy.
Basically...
I'm using a Slocket... Bad boy!
And it's NOT the same maufacturer as my MB... Bad boy again!
Slockets can kill CPU's & MB's... Bad boy for the final time!

I have to get an "approved" MB to do MORE troubleshooting, before they'll think about any RMA number.

BTW, I'm running (or trying to run), a P3-650 flip-chip in an Asus slocket, on a Soyo slot 1 board.

 

[pm]

The cache is parity protected and so you would get constant parity errors if the cache was bad. You'd boot the system and the first thing that you would see would be PARITY ERROR - then you know your CPU's cache is bad.

When a CPU dies, most of the time it literally dies and stops working completely. It would be extremely rare for a malfunctioning CPU to even get anywhere near booting into Windows. We are talking 1 in 5000 dying CPUs or maybe even more rare than this. Since CPU's statistically don't go bad very often, then this puts the odds of this being your problem somewhere in the 1 in a million thing - which doesn't mean you are wrong, Blaine, but just that you are probably wrong.

If you are getting occassional Windows errors, then I concur with my company's technical support that it's probably something else. A Windows driver issue is usually the most likely culprit. Bad memory settings, or the motherboard are two other common issues.

 

[Blain]

Here's the history of my woes.

 

[pm]

Hmmm. Blaine, let me talk to some guys at work (Intel) and I'll get back to you later this morning by email. It does sound like it's the CPU cache, but like I said you should be getting parity errors if the cache is bad.

It seems like you've looked into this pretty thoroughly. Email me the name of the rep that you talked to on the phone and I'll look into this a little more.

Email: pmahoney@mipos2.intel.com

 

[Blain]

You've Got Mail big daddy!

 

[Blain]

JUST TO UPDATE...

After testing this CPU, it turns out that some kind of compatability problem has developed between this P3-650/Asus slocket/Soyo 6VCA.

Here's the lowdown in case you or someone you know runs into a similar situation.

P3-650 WORKS with my generic slocket on the Soyo MB.
P3-550E WORKS with my Asus slocket on the Soyo MB.
P3-650 WORKS with my Asus slocket on an AOpen BX MB.
P3-550E WORKS with my generic slocket on an AOpen BX MB.
P3-650 WORKS with my generic slocket on an AOpen BX MB.
P3-550E WORKS with my generic slocket on the Soyo MB.
P3-550E WORKS with my Asus slocket on an AOpen BX MB.
P3-650 DOESN'T WORK (causes Windows registry errors), with my Asus slocket an the Soyo MB.

Now how odd is this one? :Q

 

[BradS]

What I find the most interesting is how disabling the cache can "fix" the problem.

pm - Eagerly awaiting your response.