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: Dave B who wrote (14464)	1/29/1999 5:28:00 PM
From: Dave B	Read Replies (2) \| Respond to of 93625

Hi all, Just got back from lunch. We had a great time. Sorry y'all couldn't make it as well. I took 5 pages of notes, so I'm not going to try and edit them too much -- just blast them at you stream of consciousness over the next couple of messages. At the meeting itself we had unclewest, bernard super, woodside and myself. The four of us asked about 75% of the questions. It started at 9:00, the official stuff was over by 9:10, Tate and Harmon did a presentation for 30-40 minutes (same presentation as last year, just updated with new partners, etc.). Then we had about 30-40 minutes of questions. This year there were about 30 non-company/auditor attendees (up from 3 of us last year). At lunch we had unclewest, bernard super, timothy liu, myself, and a gentleman who joined us who will be signing up shortly (watch for "Ribs", he says). Good food and good conversation -- highly recommend the Blue Chalk Cafe in Palo Alto (disclosure: I have no financial interest whatsoever in the cafe <g>). So on to the feedback, but first a thought: For any of you who thought there would be some major announcements at the meeting, that was unreasonable. They couldn't provide news to such a limited audience. You probably would have heard something before we did because they would have released it via the press. So there was nothing earth-shattering that was said (many of you will probably stop reading now <g>). I'm just going to present my notes in bullet form. If you have any questions, post them to the thread or PM me separately. To the other attendees, please feel free to correct anything you think I misrepresented. * Camino - due to non-disclosure agreements, they can't comment. If anything were to be announced about it, Tate thinks that the Intel Developers Forum might be the place * The increased engineering expenses are simply headcount. They're going to hire 10 people more than they originally planned to to meet the needs of supporting their licensees. Personnel costs make up 2/3s of their engineering expenses. * Chromatics and Cirrus Logic dropped Rambus for reasons completely unrelated to Rambus. Didn't get a chance to ask them their opinions on why it was dropped. * There are no royalties for connectors, testers, etc. The only royalties come from DRAM chips and logic chips. These other companies might be charged an engineering support fee. * SGI was the first company to actually use Rambus memory in a system * Tate said their business model was like a software company * Tate reviewed the partner list -- "got 'em all, blah, blah, blah" (no disparagement intended). No one we didn't already know about. Go to next message -> Dave B p.s Hi Jeff, missed you at lunch :-<.

To: Dave B who wrote (14464)	1/29/1999 5:43:00 PM
From: Dave B	Respond to of 93625

.... 2nd message .... * With 64M Rambus memory, the focus has been on getting the product to market. Eight companies currently have samples available. * With 128M/256M Rambus memory, the goal will be to get the cost down to 5% to 10% above 100Mhz SDRAM. This is based primarily on reducing the size of the chip. This will make RDRAM very price competitive because Rambus memory actually reduces the costs of some other components (less printed circuit board, etc.) * National Semi has committed to putting the Rambus memory directly on the processor chip. (More on this shortly). This is important because no external chipset is required (reducing system cost) and, even more importantly, Rambus is the ONLY memory that will be supported. * Rambus measures performance based on the number of bits transmitted over a single wire. In this context, they blow away the competition (DDR, SLDRAM, etc.). Right now the others might be competitive because they have wider data paths, but future chips will support wider Rambus busses as well (Rambus is organized in "channels" where each channel is 2 bytes wide). Future processors will support 2 or 3 Rambus channels. * Rambus has been improving their bus performance about 100Mhz per year for the last 3 years. DDR and SLDRAM don't even match the 500Mhz/line speed that Rambus was at 3 years ago. * While systems have been shown running RDRAM, no systems have been demoed anywhere (according to Tate) running DDRDRAM or SLDRAM. * Rambus currently has roughly 120 patents filed, of which 50 have been granted (I assume the rest are still in the process). * Digital TVs require 1.6GB/sec bandwidth. Rambus can do that today. A DDR or SLDRAM configuration would be more expensive since so many more chips would be required (e.g. an 8 byte wide data path if the chips are running at 200Mhz). * Some of the Rambus designs currently being worked on go up to 10GB/sec (more channels). * In May 1997, Rambus had 20 design wins. In January 1999 they have 70 design wins (a design win is a product that has been announced that uses Rambus inside). There are many more products that have not been announced. Next message -> Dave B p.s. Hi Jeff, we've decided we're going to say hi to you in every posting from now on <g>. You're our Rambus mascot.