Hi all; The Mindanao Rambus quote fest:
Dec 17, 1998
But there isn't a better mousetrap and the odds of one being created are pretty slim:
-- Despite years of effort and millions of dollars invested, the nearest competitor is still only 1/2 the performance.
-- The Rambus mousetrap has taken almost 10 years and a small fortune to develop. So even if a better mousetrap is in the works, it assumes that Rambus development is standing still (which it isn't).
And even if one is created, the Rambus mousetrap is protected by fundamental patents that any chip-to-chip interface technology will have to address.
The barriers to entry for "a better mousetrap" are just too huge. TheStreet.com needs to consider this point. #reply-6877526
Dec 18, 1998
So far, the rollout of Direct Rambus has been on- or ahead of schedule and the industry infrastructure (memory modules, connectors, clock chips, other suppliers) is overwhelmingly supporting Direct Rambus. The transition is going smoothly. #reply-6897021
Thomas, Something you won't find highlighted in the DDR literature: DDR is 1/4 the performance of Direct RDRAM. #reply-6897073
The DDR data sheet is lying. The data rates are not the same. In fact, it takes 4 DDR chips to perform at the level of one Direct RDRAM chip. When you're talking about the amount of money saved (bills of materials, inventory, assembly, etc.) by PC makers when choosing a specific technology, Direct RDRAM is the least expensive choice. #reply-6897142
Dec 20, 1998
For main memory, DDR is dead on arrival (for the reasons mentioned previously). But in applications where only a small amount of memory is needed, DDR can have an interim place -- as in some graphics frame buffer apps. Just to put it in perspective, though, according to folks like Dataquest, PC main memory is about 50% of the DRAM market and graphics memory is about 5%. So DDR is an interim technology for something less than 5% of the overall DRAM market. Not exciting. And certainly not a threat to Direct Rambus. ... #reply-6913901
August 26, 1999
August 26, 1999
The Rambus patents are incredibly fundamental, particularly one that was assigned in Q1/Q2 98. Remember, this is an intellectual property company. #reply-11080446
The Rambus technology (the interface, the channel, the WHOLE package) reduces system cost. People get confused because the chips themselves are a little pricier than conventional memory technologies -- but what they seem to forget is that those technologies were more expensive when they were relatively new, also. But if you keep your eye on the ball (that Rambus improves the performance and lowers the cost to create electronics (computers, communications systems and consumer electronics), you see that there isn't any real alternative to this technology. #reply-11080755
Of course the constraint is bandwidth... for the applications Rambus targets. Eventually, every application will hit the wall (game machines went first, remember). There's lots of money still to be made in the applications in which the company is currently strong. #reply-11081431
August 29, 1999
... As for DDR, I think I'm pretty well on record with my opinion on that (might work for some graphics apps, but it dies in main-memory configurations, no matter what the "data sheets" say). ... #reply-11098632
-- Carl |
