Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Rambus (RMBS) - Eagle or Penguin -- Ignore unavailable to you. Want to Upgrade?

To: Tenchusatsu who wrote (35234)	11/26/1999 6:27:00 AM
From: Bilow	Read Replies (2) \| Respond to of 93625

Hi Tenchusatsu; Re IBM's 5000 pin server. Thanks for verifying that the design is real. Sure it's hard, but bleeding edge technology eventually becomes standard issue, especially when it solves important industry problems. My mention of that part is to illustrate the trend in package types, and it is very clear. More pins. A lot more pins. Current DRAMs are x16. The next generation is x32. Re "the uni-directional pin runs at 1.6 Gbit/sec" type of interconnect. I think that the stuff is on the right route to the future. But it isn't a general purpose DRAM interface. Unidirectional and point to point connections are a lot more robust than the Rambus solution which is multidirectional and multidriver, and is what is used on general purpose DRAM interfaces. Sure Timna has an RDRAM interface, but that decision was started before RDRAM went TU, and you have to wonder if Intel is working on an SDRAM version of the same chip. Intel is undoubtedly going to ship most of the RDRAM controllers with SDRAM translators, just like they are shipping the Camino chips, as reported repeatedly (by former RMBS bulls) on this thread. The thing to watch are design wins after the summer 1999 debacle, not design wins from 1998 and before. Re "we don't know whether Nintendo is using DDR". This was the subject of an EE-Times article that has been linked to at least once on this thread. I suppose I could edit in the link: The DRAM will be "an application-specific memory" with a data transfer rate of 3.2 Gbytes/ second-presumably, based on the emerging DDR II specification-and dedicated for use in Dolphin, said Keiichi Shimakura, associate senior vice president of NEC. techweb.com "Application specific" means not a standard DRAM. Why should they pay royalties to Rambus if they're not going to use a standard memory? Not much reason. If Nintendo were using Rambus, I think that they would have said so. Instead they said: High Speed DRAM Technology nintendo.com Re "I can't even get DDR yet, much less DDR-II." You can get DDR anytime you want to by ordering it. I've had parts in my hand as long ago as March. It is in production from most of the memory makers. They aren't selling to retail, but they are certainly selling to me. And it is design engineers that count, not the retail PC trade. Design engineers are what determine memory usage in the future, and they are who Rambus sells to. Sales of memory in 2001 are largely determined by design wins in 1999. Rambus died in mid 1999. You won't see product from early 1999 design wins until 2000, so you will see new products coming out that use Rambus. But that is a lagging indicator. The leading indicator is what design engineers are doing right now, and they know that Rambus is dead. Nintendo is just the tip of the iceberg. The vast majority of memory design wins go unmentioned in the press. The loss of the most important Rambus customer should indicate to even the most hopeful long that the technology is dead. You'll see the DDR-II when the design wins it is picking up start shipping. Intel's support is melting away. No reasonable observer could deny that Intel's support has decreased since a year ago, the only possible disagreement is how low that support will fall in the future. Neither can anyone realistically deny that direct RDRAM has seen delay after delay. -- Carl