To: Dave B who wrote (77536 ) 8/16/2001 2:21:24 AM From: Bilow Respond to of 93625 Hi Dave B; Re: "Again, there was a lot of animosity and so on directed at Rambus before they went after the SDRAM and DDR royalties, so that's not the issue. " This is true. Rambus' management was always incompetent, no one is trying to say that going after SDRAM and DDR was their first error. My issue is that Rambus' management used sleazy business tactics and had lousy technology long before they went after SDRAM and DDR royalties . The 4i initiative is a classic example of Rambus incompetence; they accidentally added 25% to the die size of RDRAM without improving performance, now industry is removing it. In fact, the only reason they went after SDRAM and DDR is because they were incompetent. First, they got a deal with Intel that would have pushed RDRAM into the mainstream except that the technology was so bad (i.e. not cost effective against the competition.) Then, after their incompetence led them to a failed deal with Intel, they went to "Plan B" which was to incompetently threaten the industry with their SDRAM and DDR patents. Now they're in trouble. Re: "As for trading patents, sure it's a great idea, but are you saying that only those in the "industry" are allowed to come up with ideas that affect the industry? " No, unfortunately. What you are illustrating (and Rambus' actions illustrate as well), is that in the modern world, IP is the soft underbelly of technology. But we never got to find out what Rambus' IP would have been judged, in a court of law, to be worth. The bears have many times noted that 10s of thousands of patents apply to DRAM, so Rambus' share of these is minuscule. Sure we've got statements from SI experts that Rambus can legally charge whatever they want for their IP, but hey, every one of the "experts" who said that was dead wrong on the results of the Virginia case, both the infringement and the fraud. Yeah, it's possible to make incredible money as an IP firm, look at QCOM. But Rambus' IP isn't the same as QCOM's IP, they're Apples and Oranges. QCOM's IP is more like SDRAM, if you want to play, you have to use it, and of course that is exactly why Rambus tried to patent SDRAM. As far as analyzing the theoretical (as opposed to practical) advantages of RDRAM over DDR, give it up. None of you have the job experience to do that. It should be enough to note that RDRAM is dead. If it were this brilliant technology, Intel wouldn't be phasing it out, and Samsung wouldn't be cutting over to DDR production. If you want to believe that the Rambus ramp continues on, hey, go buy some more Rambus shares. But over the next few days / weeks, Wall Street is gonna figure out that Samsung is ramping down on RDRAM, I think it would be a good idea to not be holding RMBS when they do. I've already been over with you guys what the future holds for DRAM. It's obviously going to a combination of embedded and MCM type packaging. There's not much room for Rambus there. The problem with Rambus for the future, is that the latency, the complexity, and the bandwidth of Rambus, (as well as the royalties) are each lousy compared to the competing technologies. Rambus was only a brilliant technology in the ivory tower dream world of a few (now very wealthy) theoretical scholars, it was never practical for mainstream use. -- Carl