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: richard surckla who wrote (36185)	1/3/2000 11:01:00 PM
From: Bilow	Read Replies (2) \| Respond to of 93625

Hi richard surckla; Still dead. -- Carl

To: richard surckla who wrote (36185)	1/3/2000 11:02:00 PM
From: Jdaasoc	Respond to of 93625

richard: "Rambus in PCs may be in a shambles now, but it's still to early to count Direct RDRAM out entirely. MicroDesign Resources analyst Peter Glaskowsky sees system engineering problems with Direct Rambus as inevitable - but not insurmountable." Article seems to be primarily about RDRAM in laptops but is another confirmation of the backing off of Intel regarding widespread use of RDRAM in desktops. eb-asia.com Taiwan, Direct Rambus DRAMs Don't count Direct Rambus out yet As Intel is reportedly engaged in consensus building with major DRAM makers on the next-generation memory in PCs, the world's largest chipmaker has distanced itself from partner Rambus Technology. That change in position partly reflects delays and costly mistakes during 1999 surrounding Intel's Camino chipset, once expected to make Direct Rambus the mainstream memory in desktop PCs starting next year. In addition, Electronic Business Asia more recently broke news internationally about the cancellation of the Greendale chipset, which would have enabled Direct Rambus in notebook PCs. Intel's grand scheme to implement Rambus in PCs may be in a shambles now, but it's still to early to count Direct RDRAM out entirely. MicroDesign Resources analyst Peter Glaskowsky sees system engineering problems with Direct Rambus as inevitable - but not insurmountable. In the long term, according to Glaskowsky, the industry has little choice but to move to more demanding memory technologies such as Direct RDRAM. "Increasing the data-transfer rate by any means will raise these issues. If OEMs don't want to increase the performance of their systems, that's fine. They can keep using PC 100 SDRAM. If they want a slight increase, they can go to PC 133, which is slightly more sensitive to board design. If they want more of a speedup, they'll need to do more work. There will never be a way to double the speed of memory without solving issues like these." Lead times for product development and design validation of Direct Rambus PCs will undoubtedly be extended. Nevertheless, a spokesperson for one notebook maker in Taipei says, "Rambus for notebooks is only on hold. Rambus-equipped notebooks will probably not start to appear until the end of next year or early 2001." The source also indicated that the Solano 2M, widely seen as Intel's "plan B" chipset following the cancellation of Greendale, could also support Rambus. According to the source in Taipei, "The Solano 2M itself supports PC 133 memory, but theoretically it can still support Rambus. This might mean one or two design tweaks, though. We might see 2M1 or 2M2 parts, for example, before we finally get a satisfactory Rambus solution." According to analyst Glaskowsky, "The problem is that the [Rambus] interface has many fewer pins, so its power consumption is more concentrated. In fact, the interface power can be concentrated in a single device. This is why desktop RIMMs need heat-spreader plates. Mobile designs have less tolerance for that sort of thing, and historically there's been little or no cooling-air flow through the volume occupied by memory." Notebook makers in Taiwan have encountered major difficulties in maintaining signal integrity when they attempt to move portable machines to the tighter design definitions demanded by high-bandwidth memory. One source at a leading Taiwan notebook manufacturer explains that signal noise may not indicate problems with Intel's Greendale chipset itself. "Satisfactory Rambus implementations could pose more a [system] engineering problem than a chipset problem. Parameters for re-design become much tighter. Engineers are working within tight constraints. There's less room to maneuver. Rambus makes layout and design engineering very critical. Rambus adds a higher degree of engineering complexity." Glaskowsky thinks that for now, Taiwan notebook makers will play a careful game, but they will get to Rambus or an equivalent high-speed memory sooner or later. "These companies are being patient and conservative, extending their SDRAM-based core-logic designs as far as they'll go, probably through some form of evolved DDR memory that will take them to 2002 or so. After that, they will have to switch to Direct RDRAM or some hypothetical alternative. I believe when they're ready, they'll decide it's easier to pay a little for Direct RDRAM than pay a lot for some untried technology with similar challenges."