Chas, I agree with your point that if all these players suddenly see the light on RDRAM that will affect the supply of other types. How much the availability of RDRAM shifts OEM demand for other stuff - who knows? I'm wondering whether there will be a bit of "sticker shock" when the OEM's see the prices of RDRAM pieces, but then again if they're going to be going into boxes for the high-end segment, maybe that's less of a concern or even a non-issue. That price premium is what makes things tricky. Even though Samsung is projecting 50% of sales to be RDRAM, that premium makes it tough to backout how much of their % of total output in megabits will be RDRAM.
And also, I thought it was interesting that the article talking about 72-bit parts, but seemed to be mute on the 128's.
MU seems to be willing to jump right into the 128's.
So by jumping straight into the 128's, on the one hand maybe you sacrifice some of the early premiums on RDRAM pieces, but then again, if everyone else is out shifting resources away from 64 PC100 production, and MU isn't (while still maintaining their 128 development) and what you're pointing out happens - that ends up being good strategy on MU's part (assuming their 128 development doesn't take a meaningful bite out of their own non-RDRAM production).
Asking how much the fragmentation complicates the planning - that's a great question.
Good trading,
Tom |
