Rambus, Micron, Ms. Burstiner, ...
After reading what I thought has been some biased reporting by Ms. Burstiner, I was actually impressed with the article posted on TheStreet.com. Seemed reasonably balanced, at least to me (note: I'm very long on Rambus).
And rather than discuss the long-term issues, since I think it's fairly clear that Rambus's technology wins in the long-term, let's discuss the short- and intermediate term.
So what this comes down to is whether Micron will get what it wants (a frontside 133MHz bus and a 133MHz chipset to SDRAM) or not. It's difficult to believe that Intel would do this unless they saw that Rambus/the DRAM manufacturers were going to ramp up slowly enough such that they were forced into such a move through a threat to Intel's higher-speed processor roll-out later in the year.
There are several issues here:
1) How difficult is it for Intel to implement a 133MHz
front-side bus and chip-set for SDRAM? If I were the
R&D manager of Camino, I think I'd scream about having
to do this after already being in charge of a late
project that was my fault for being late. Anyone got
any ideas about just how difficult this would be technically
for Intel?
2) Does anyone have any insight into the progress being made
by the DRAM manufacturers implementing Rambus, other than
the excellent press releases that REH, Uncle, and others
have posted? Mr. Hume in particular tends to posit that
he believes that DRAM manufacturers are slow in their
progress -- anyone see proof of that?
One note on #2: I don't see that Micron's position as evidence that DRAM manufacturers are slow on the Rambus ramp. The reason is simple: if I were Micron, what I'd want from Intel would be for Intel to take a position that would maximize my profit. For Micron, that's simple: they want Intel to do more R&D/marketing such that an overlapped 133MHz SDRAM and Rambus release is performed so that Micron can milk their existing DRAM technology (just upticked in speed) while working out any potential kinks to Rambus adoption that might be local to Micron. This is particularly important for Micron since the company has benefited lately from many of its competitors shutting down memory fabs.
Anyone got any opinions, with backup facts/theories, of course, with respect to this?
Thanks! |
