mb, re: rambus
I can understand the negative sentiment on the board, even considering joining the "Poot" club.
The problem I have, it's a bet that Intel won't get what it wants and perhaps needs. By 2001, entry level will probably be ~750mHz with 1 gHz on the high end. And at equivalent prices to today's mix. These chips will require a highly engineered electrical and mechanical bus. I think we're talking matched transmission line properties, even through the connector. There's also the issue of intellectual property as a barrier to competition. Now that the FTC's off their back, it look's like Intel can use all of it's legal prowess to maintain control of Rambus issues.
And I'm not sure the market will have much influence. With powerful relationships, (and power over relationships), I think that most providers will produce what Intel wants. A lot is made of costs, but when I read that Rambus is only 35% more expensive than SDRAM, is that really a major issue in todays memory market?
I wonder if it isn't a mistake to dwell on what will happen over the next few quarters. Even if PC133 becomes successful, Intel will drive for Rambus in systems. And with time the need may become more apparent. Systems will get faster, they always have, even if we don't perceive the need for all this performance. (And across the price segments). But as we know, system performance is much more than the speed of a CPU. That could be a very legitimate reason that Intel is pouring heavy resources into the product.
Doesn't RMBS valuation has a lot more to do with Intel than with Rambus the company. Is that how the market looks at it?
These are all just my opinions, of course.
Gene |
