One of my favorite quotes from the 19th century master sleuth himself goes like this: “It is an old axiom of mine that when you have excluded the impossible, whatever remains, however improbable, must be the truth” (The Adventure of the Beryl Coronet, pg. 315). Would that I had recalled his words earlier when working my way through a recent troubleshooting adventure (read all about it in my latest ViztaView blog entitled “Test System Woes Finally Solved, But Not Without a Final Fillup of Loathing and Despair“).
To compress this epic troubleshooting adventure into as few words as possible, when trying to figure out why Windows 7 Professional x64 wouldn’t install on a particular test machine, my reluctance to consider the CPU as the possible culprit caused me to waste ungodly amounts of time trying to fix (and even replace) other stuff that wasn’t broken or misbehaving. It turns out my engineering sample of the Intel QX9650 processor (which samples, by the way, are neither supported nor warranted to be defect-free by Intel: they give ‘em way for reviews and analyses and you gets what you pays for them) simply won’t complete the “Expanding files…” phase of the Windows 7 install process, which follows immediately after “Copying files…” right at the very beginning of the install process.
I’m actually writing a story about this for ITExpertVoice.com, as a combination advanced troubleshooting tutorial plus a meditation on the nature and essence of systematic troubleshooting. So naturally, my editor at the site asked me to get a comment from Intel to make sure this leading silicon foundry didn’t get blindsided by my report. First and only official response I got was “Intel modern processors work on both 32 and 64-bit versions of Windows 7.” I’d have to agree with this statement and furthermore I believe that my particular borked QX9650 is an anomaly and not in any way representative of Intel’s architectures, products, and capabilities.
In off the record discussions after that, however, things got really interesting. Basically, the Intel rep refused to believe that I was reporting a genuine phenomenon and that my troubleshooting methods must be flawed or incomplete for this to occur at all. Nevertheless, I can repeat this anomalous behavior at will, and the only single factor that either causes this behavior to occur, or that makes it vanish, is the presence (problem occurs) or absence (problem disappears) of this particular QX9650 CPU. If everything else either stays the same or is different, and the only element that controls whether or not the problem manifests is the CPU, logic and reason (thanks, Sherlock!) tell us that no matter how much we may not *WANT* to believe the cause, it is indeed the truth.
It’s stuff like this that makes my working life so much fun, and such a treat to keep whaling away at. Sometimes I feel like the luckiest guy in the world, but only after I recover from feeling occasionally accursed that things don’t always work the way they should!