Original Pentium FDIV flaw e-mail [1994] | Hacker News Reader

Original Pentium FDIV flaw e-mail [1994]

trnicely.net

67 pointsjonah17 comments

I vividly remember this. I grew up in Lynchburg and one of my best friends' fathers was a professor at Lynchburg College. At the time this happened I was in my senior year of high school and completely immersed in the school's new computer lab (Wordperfect 5, Mosaic, Matlab, Pascal -- woohoo!). I went to hear Dr Nicely speak about the flaw and thought it was super-cool that this was happening in my [really boring] hometown -- thanks for dredging up history. :)

Intel's CEO, Andy Grove later wrote in his book "Only the Paranoid Can Survive" that the crisis had marked Intel's transition from being run-of-the-mill B2B operation to major B2C brand.

Before the crisis Intel viewed itself as just yet another supplier of microchips to powerful OEM's, a pure business to business operation that had computer manufacturers as their customers. They genuinely believed that it's up to end-product manufactures to handle user's compaints and then settle any issues with Intel account managers in the privacy of company meeting rooms.

What Intel didn't realise is that their prior advertising campaign that was hammering "Intel Inside" into consumers' minds and re-enforcing the belief that a PC powered by Intel microprocessor is superior to any other PC was incredibly successful. Intel became a consumer brand, but the company didn't have the structure, nor the mindset to serve the end-consumers directly!

When the denial was over and it became apparent what the actual problem is - not the microprocessor flaw, but the company's perception of itself - Intel moved with energy, quickly and decisively: hired hundreds extra staff, changed reporting lines, created an entirely new end-consumer servicing structure - all within weeks.

I thoroughly recommend reading Andy Grove's book for the Intel's side of the story.

http://www.intel.com/pressroom/kits/bios/grove/paranoid.htm http://www.amazon.com/Only-Paranoid-Survive-Andrew-Grove/dp/...

I remember this mostly because my boss at the time referred to Pentiums in the plural as "Pentia". "Pentia". Let that one sink in. He also was a chain smoker who would carry his ashtray and smokes into the machine room and then wonder why the big Power Challenge had mysterious SCSI bus problems. Good times.

I'm not really picking up why it's so reprehensible for someone to jokingly pluralize Pentiums as Pentia. Can you explain?

It wasn't a joke.

It's odd, but someone actually downvoted me for getting you to clarify that. Haters gonna hate, I guess.

And it wasn't incorrect either. -um is the ending for neutrums (actually neutra) and -a is the plural thereof.

True, except in that Pentium is an English neologism, and thus is pluralized like an English noun. Don't get me started on the plurals of octopus or virus, please.

Well, if my Latin knowledge hasn't abandoned me -um is the ending for masculine words whereas -a is used for feminine words. I suppose the proper plural of Pentium would be Pentii :p

-um would generally be a neutral-gendered noun in Latin. (You're thinking of -us as a masculine ending). Thus, Pentia is the correct pluralization.

However, Pentium was at best a pseudo-Latin brand name to begin with, and Pentiums is likely the least wrong pluralization in English.

Yeah, you're right now that I think about it (the example that reminded me is pluralizing walrus to walri). Personally I do this sort of thing just for personal amusement (as I imagine this persons boss did) rather then actual correctness.

Here is the timeline:

http://www.emery.com/1e/pentium.htm

USENET grew one pant-size that autumn.

Did it ever, and this parody stuck in my mind:

Open the pod bay door, please, Hal... Hal, do you read me?

    Affirmative, Dave. I read you.

Then open the pod bay doors, HAL.

    I'm sorry, Dave. I'm afraid I can't do that. I know that you and Frank were planning to disconnect me.

Where the hell did you get that idea, HAL?

    Although you took very thorough precautions to make sure I couldn't hear you, Dave. I could read your e-mail. I know you consider me unreliable because I use a Pentium. I'm willing to kill you, Dave, just like I killed the other 3.792 crew members.

(etc.)

What sticks in my mind the most about this was the estimates on the part of PC manufacturers, of the estimates of the probability of a spreadsheet user, for example, running across this bug. Thinking through the problem carefully, there was no way to calculate such a probability.

I was oddly tickled to see that acavax.lynchburg.edu is now serving http, with an index page whose last modified date is not quite 10 years ago.

It was one of the best things that happened to Intel. After some reluctance, they proved as a company users can trust in, as they allowed to return their CPUs in exchange for a non-broken one. This trust into Intel was worth much, much more then the cost of the refund policy.

It took Intel weeks to realize that they had a genuine PR disaster on their hands, and if it had taken them much longer to agree to replace any defective processor, the Pentium brand would have been ruined. They learned their lesson and wholeheartedly did everything they could to smooth things over, but only after a month and a half of making things much worse. By the time Intel started dealing with the situation intelligently, they were facing several lawsuits and even the mainstream media was making jokes at the expense of their flagship product.

Reply on news.ycombinator.com