James,
Thought I'd weigh in on Edelstone's comments. Possibly he meant the following...
The biggest "problem" with the Internet right now isn't getting text and graphics from the computers memory into the processor (or video memory). The biggest problem is getting data down the pipe to the user. Someone accessing the internet at 56K wouldn't see a significant difference in performance if they were suddenly given a Rambus-based machine. They would, however, see a difference with ADSL or a cable modem. So Rambus doesn't suddenly make the Internet more exciting. It will, however, be an enabler over time, helping improve multimedia usage, etc (as will faster processors, bigger disk drives, yada, yada, yada). Keep in mind that webmasters won't start suddenly designing huge multimedia pages because: a) there will still be pipe limitations, and b) there will still be many more machines with non-Rambus memory, even if it does help.
My father bought one of those $399 computers, and for what he does on the Internet, Rambus would be a waste of money.
On the other hand, the segments that would benefit greatly from Rambus memory are memory-intensive applications like multimedia, database management, routers & switches, and areas where lots of data needs to be moved INSIDE the computer, not TO the computer. So acceptance in these areas is much more critical.
JMHO. I'm just trying to interpret his statement. Internet usage will continue to grow dramatically as low-cost machines continue penetrating more and more homes. Rambus in those machines won't drive any additional business or benefit for the Internet for a while (years, at least). Over time, Rambus will appear in those machines as the cost differential comes down and everyone stops manufacturing SDRAMS, but the biggest benefit initially is in the other areas.
Dave B |
