To: Don Green who wrote (67240 ) 3/6/2001 6:04:20 PM From: Bilow Read Replies (1) | Respond to of 93625 Hi Don Green; You really should give up on trying to find my factual errors. If I state it, it is very likely true. If I state it repeatedly, you had better be sure that I can prove it. I've been here at least since Oct 1997. That's what, 3 years, 5 months, yeah that counts as "years": Bilow, October 9, 1997:Article from EETimes: << Though SGI was one of the earliest backers of Rambus Inc.'s memory architecture, it was driven by cost considerations to industry-standard 100-MHz synchronous DRAMs. At the same time, SGI has set up agreements with several companies to provide it with special packaging and modules. "We found that with commodity DRAMs we were able to achieve the performance we needed," Trimuschat said. "We're experts in designing systems, not in providing DRAMs." >> EE Times:techweb.com -- Carl #reply-2405257 By the way, did SGI, "one of the earliest backers of Rambus Inc.'s memory architecture", ever come back and use Rambus again in the 3+ years since then??? Nope. No company that ever used a Rambus generation ever came back and used it in their next generation product. Instead, they immediately began development of alternative memory interfaces for their succeeding generations and folded them into production about as fast as they became available. Big companies have a lot of inertia, so if you're looking for a quick move you aren't going to find it here. But the trend is clear, RDRAM has been rejected by industry. Back when Intel had 75% of the chipset market and said that they'd never support PC133 or DDR it was possible to believe that RDRAM had a chance, just due to that much market share for Intel chipsets. But now even they are combing the Rambus nits out of their hair, and they lost most of their chipset market share to companies that weren't married to Rambus. -- Carl