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: ekid who wrote (66814)	2/28/2001 7:04:50 PM
From: Bilow	Read Replies (3) \| Respond to of 93625

Hi ekid; Re Samsung memory problems: eetimes.com This was after Samsung spent big bucks trying to get RDRAM to work. The R&D costs were huge, particularly compared to other memory types. You have to read between the lines, but even mom and pop types probably understand what "idle" means to a memory maker: "One day, I woke up and found that the once-idle* RDRAM line is fully running like a fire hose and only able to produce 6 million to 7 million 128Mbit equivalent RDRAM chips,"* electronicnews.com Samsung is making good money on RDRAM right now, but they don't know how long before the market for RDRAM peaks and then declines. In the past, they built up their RDRAM lines for an Intel ramp up that never happened, and then found themselves with idle lines. That's expensive, so now Samsung doesn't want to convert more lines from SDRAM to RDRAM. Now do you understand why it is that Intel has to pay Samsung to expand their RDRAM lines? The rest of the industry isn't fooled by the malarky about Intel's future memory plans given out this past week. For that matter, the rest of the industry wasn't fooled when Intel said the same thing (and invested the same money in the memory makers) over the PC133 vs RDRAM question of 18 months ago. The issue is what chipsets does Intel have to offer, and because of management error at Intel, they pretty much only have RDRAM chipsets. Until Intel can replace those chipsets with SDRAM and DDR chipsets (like the i815 replaced the i820), Intel is stuck supporting RDRAM. If they announced tomorrow that RDRAM was dead, dead, dead, they would have more difficulty selling P4s than they already do. So why would I (or you) expect them to? Of course they're not going to say that RDRAM is dead. You don't bet a multi billion dollar company on the ability of a tiny Taiwanese chipset maker (or three) to deliver on time. In addition, a multi billion dollar company can't pump out a DDR chipset quickly. So Intel has to assume that there will be no DDR chipsets for P4, and therefore they have to make sure that there is as much RDRAM available as possible. Thus, Intel has to make investments in the memory makers. Watch for Intel's position on RDRAM to begin to reverse if VIA or ALi start sampling a P4 chipset, as is inevitable. This whole thing is in complete distinction to the SDRAM memory conversion, which required absolutely no investments from Intel. In addition, other chipset designers supported SDRAM from the get go. With RDRAM, there are still no design wins for Intel compatible PCs outside of Intel. On the other hand, there are many design wins for DDR chipsets for Intel compatible PCs. Count them here: users.erols.com Let's look at the sequence: (1) Intel says that after PC100, the next memory standard is RDRAM, and that PC133 won't work, and they will not produce a chipset for it. (2) Intel announces a chipset for PC133. (3) Intel says that after PC133, the next memory standard is RDRAM, and that DDR is a bad idead and won't work, and that they are not interested in producing a chipset for it, except maybe in the server industry. (4) Intel announces that they are working on a desktop DDR chipset. (5) Now Intel says that DDR is only a temporary memory standard and that the next real standard is RDRAM. At this point, after having been lied to repeatedly by Intel, you still believe them? Now I begin to see why *[Insert name of most hated political leader here].* Clinton still had so many supporters. You guys believe what you're told. The industry knows RDRAM is dead, and so does Samsung, for that matter. Press releases are not indications of the beliefs of companies. Actions are much more indicative, and the fact is that Samsung is charging Intel to expand production of RDRAM. Niche product. So do please answer the question: If Samsung is making good money on RDRAM, and they think that RDRAM chips are the inevitable wave of the future, why is Intel having to give them money in return for promises to raise production of the things? Also, why is it that VIA, ALi, Micron, SiS, ServerWorks, Nvidia, etc., aren't designing RDRAM chipsets? And where is Samsung's guarantee of pricing parity at the RIMM level? -- Carl