With regard to RAMBUS
Intel may be paranoid (which is good) but with respect to RAMBUS they have been foolish (which is bad). They made a bad bet, and simultaneously executed a bad safety net (the MTH).
Bad decisions. Bad engineering. Bad Marketing. They left nothing to chance.
I love this company and it has made me a small fortune. But this may rank up there with the watch business, the iAPX432 and the FDIV error as proof that paranoia is still not a substitute for good judgement. Paranoia protects you from external threats. Something else is required to protect you from yourself. Usually that something is the marketing department, which might have said something like, "Why don't we let the market decide whether the benefits of Rambus are sufficient to warrant a premium price. Why don't we offer BOTH a RAMBUS solution (that works) and a non RAMBUS solution (that works). Offering either only RAMBUS solutions or solutions that don't work could be suicidal. Or a the very least a serious, serious black eye that will piss off our customers the boxmakers and drive them into the arms of our rival(s). This is a very, very dangerous decision and it is probably going to hurt us. Big."
Normally, Intel is very good at recovering from a screwup, but Intel, in this case, is making a fairly poor situation hideous. Unless magically, Rambus volume explodes up, prices come down to a modest premium over standard ram and AMD develops production hiccups of substantial proportions (all fairly unlikely) Intel has minimized what should be a great opportunity to ice their position as the dominent microprocessor vendor.
Not too many rational explanations for this blunder. It will be studied in the top business schools for the next few years under the category DON'T DO THIS. The good news is that Intel is strong enough to ride out and overcome even this blunder. The bad news is that they didn't have to get into this position in the first place. It's completely self inflicted. The only saving grace I can think of is that there are standard ram versions of some of the chips that are experiencing problems back in a skunk works that are being tested for producibility right now. But I don't think so.
And I like this company and think highly of its executives. I've been an investor since 1983. But a screwup is a screwup and this one's a beaut.
Good investing,
Burt |
