Carl says:
In the long run, Rambus is very doomed, due to inevitable technological change. Probably within the next 4 years, most DRAM is going to be sucked onto the processor. Embedded will still be way too expensive, but other techniques for combining DRAM and logic into a shared package will proliferate...
In the long run, the question is "Is Rambus a one trick pony?" If so, then they better get themselves acquired PDQ (which brings up the inevitable "by whom?" question). However, if these guys are a legitimate, long term player, then they would seem to be the leading candidate to develop the IP and other techniques you postulate.
When Rambus technologists sit around the ol' cappucino machine, I wonder where they think they want to be when they grow up? What sort of acquisitions do they talk about? If they acquired the IP for a graphics device, for instance, what cool stuff could they build? What if they acquired VIA -- what would that do? What markets outside of memory that has bandwidth problems could they enter? Or should they move their business mode to marketing other types of IP altogether?
From my perspective, it's not about Rambus in 2000 or 2001 or even 2002. It's about whether they can leverage this huge opportunity they have today into something considerably bigger. Of course, it would be nice if we could get today under our belts!
Regards,
Mark |
