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: Jdaasoc who wrote (28676)	9/5/1999 2:39:00 AM
From: Bilow	Read Replies (4) \| Respond to of 93625

Hi Jdassoc; Quite a lot of responses besides yours. Thought I'd put some comments here in just one place. Re: I must have missed class the day they discussed Yanomano. They are the hunter-gatherers down in the Amazon forest. One of the books about them, The Fierce People, is pretty much required reading in freshman anthropology courses. I found the book, by Napoleon Chagnon (sp?) interesting reading when I took Anthro 100. Since then, I bought another book on them, and found it hard to put down too. Both these books were on the NYT best sellers list at one time or another, which no book on memory design has ever (or will ever) made big sales. One that was close was Tracy Kidder's (sp?) Soul of a New Machine, which was about Data General (formerly DGN). A great introduction to CPU design for the layman, by the way, with no technical knowledge required. Regarding AMD and DDR, KZNerd posted that the AMD Athlon sales would be constrained by lack of DDR SDRAM. This is not true. The current version of the AMD chip set doesn't even support DDR. They may support it in the future. Here is an engineering overview on the AMD Athlon chip set, which includes its memory controller, from the AMD site: (pdf) amd.com Some technical notes of interest in the above paper. (1) Note that the system controller is in a 492-pin BGA package. I earlier referenced the fact that modern packaging is providing more and more cheap pins per package, thus reducing the requirement that high bandwidth be provided per pin. Reduced pin count is rambus's only essential reason for existence. (2) The memory supported is 100MHz PC-100 revision 1.0 SDRAM DIMMs. (3) The Athlon includes three 200MHz data paths. Each data path is 8 bytes wide, giving a total peak bandwidth of 5GB/sec. The high bandwidth spec on this SDRAM based processor should be a clue to those of you who think that a high speed processor requires a high bandwidth out of each memory chip, and therefore that rambus has a lock on the future. This sort of reasoning just doesn't apply to memory design. Similar reasoning would suggest: Cars use gasoline. Gasoline has to get to the engine to burn. If the engine doesn't get enough fuel, the car will be slower. Therefore, replacing the fuel lines on a car with lines 10 times wider will speed the car up by ten times. By the way, DDR SDRAMs are now available with reasonable lead times, but you have to buy them in such huge quantity (something like 280 at a time) that it is difficult for the small guy to get a hold of any. They really aren't making it easy for small users yet, and of course there are no retail channels with them. KM, I have had no position in RMBS for a long, long time, and don't plan on having any in the future, though it might be tempting to short if INTC announces support for DDR. John Walliker seems to feel that Rambus will give the user a more inherently reliable memory product. This is silly, given that the rambus technology has already demonstrated that it is having manufacturing trouble. I work with high speed digital interfaces every day, and am quite sure that it is going to be a lot harder to get technology to run at 800MHz than at 250MHz. The delays in production bear this out. To suggest anything else is at best silly. Later, when your rambus machine is available, they are still going to suggest that you not swap chips (which might be impossible to do for the average user, anyway). Dave B asks me how I know that INTC engineers were against rambus. Like any other profession, EEs have their ways of communicating. I have no intention of getting anyone in trouble with their boss, so I will not say more, except to say that I do know. Regardin the rambus bears thread, such peripheral threads get cleaned out by SI every once in a while. If I posted over there my post might disappear by the time it was time for me to say "I told you so." I will post a reminder when the time comes, but that could be a long time from now. Companies that take a highly complicated technical issue to the public in order to hype a stock price are able to keep their stock price high for years after the promises are broken and the dream is dead and gone. Look at PRST, for instance. That is why I am not short this stock. -- Carl