> 10,000 was a number that I actually read at the time. Perhaps, that was how many actually reached end users. No
> matter, the effect is still the same, you have no disagreement with me.
Just as an aside, I believe that the "10,000" in "under 10,000" is dramatically exaggerated. Dell hadn't shipped a single 1.13GHz PC, before or after the recall (or during). IBM, I think, was the only vendor taking offers and shipping product.
When the 1GHz Pentium III was preannounced in Q1, I went and phoned the various OEMs to ask them. The 1.00GHz Pentium III was in initially going to IBM, HP, and Dell. As with the later-on 1.13, Dell had not taken any orders for 1.00GHz systems. IBM and HP had, though, and I tried to get information from the two of them.
I learned that IBM was to receive 39 processors in their first shipment, which was to occur the week after the press release for the product. I also learned that HP estimated shipment numbers that month of about a hundred cpus a week, and the implication here was that a typical release of the previous year led to shipments of more like 200 to 300 cpus a week. Anyway, I did some rough guesstimates based on this and came to the conclusion that this would likely mean that less than one thousand 1.00GHz PIII systems would appear by the time that press release was one month old.
The reason why I doubt that the number of 1.13GHz PIIIs numbered anywhere in the thousands, much less near ten thousand, was because fewer OEMs were offering 1.13GHz PIIIs at release than were offering 1.00GHz PIIIs at release. Not counting Dell, which was not taking offers during the entire month after either release, two OEMs were making 1.00GHz systems at the release of the 1.00, and only one OEM was making 1.13GHz systems at the release of the 1.13.
Mind you, this reasoning takes a bit of a leap of faith (you have to assume that IBM, the one remaining OEM, wasn't getting thirty to one hundred times as many processors as they got for the 1.00GHz PIII), but I personally think that "one thousand" is much closer than "ten thousand" for the number of 1.13GHz PIIIs that needed to be recalled from OEMs.
-JC
source information:
jc-news.com (note: the cnbc article seems to have disappeared, which is annoying)
jc-news.com
PS: Just noticed that "Pentium" isn't in SI's spell checker |
