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: Steve Lee who wrote (63271)	12/21/2000 1:51:57 PM
From: Ali Chen	Read Replies (1) \| Respond to of 93625

<what do you think is the average number of chips on a DIMM> And who cares? It was proven by the market that the alleged flexibility in number of chips on a RIMM is a baloney. As well as many other advertized "advantages" of RDRAM: Message 14410527 Message 14411792

To: Steve Lee who wrote (63271)	12/21/2000 3:05:32 PM
From: Bilow	Read Replies (3) \| Respond to of 93625

Hi Steve Lee; Re: "what do you think is the average number of chips on a DIMM out of all the SDRAM DIMMS sold worldwide in the past 12 months." I'd say about 8. Funny, that would be about what I would guess was the average number of chips on a RIMM out of all the RDRAM RIMMs sold worldwide in the past 12 months too. What you thought was that there was some essential difference between RDRAM and SDRAM in terms of how many chips would be put on a typical module. This comes down to the old "granularity" argument. RDRAM would have an advantage over SDRAM because the parts have lower granularity. There were two errors in the argument. The first was the assumption that SDRAM couldn't be adapted to have a lower granularity (in other words to add pins). The second was the assumption that the granularity currently offered by SDRAM (or to be offered in the near future) was too coarse to compete with SDRAM. Neither turned out to be the case. Heck, you may as well come clean with regard to your statement. Here it is again: "You need 8 SDRAM or DDR-RAM chips on a module. Rambus doesn't have this requirement. Hence you need a lot more DRAM chips for non RDRAM memory to get the equivalent memory capacity." #reply-13111432 The basic fact is that when you reported the above bit of misinformation, you had no direct knowledge of the facts. You were probably simply repeating something that you'd heard on Yahoo, something that sounded correct and seemed like an important point. It wasn't your fault that your post was in error, you're just another 20-something, with no time in the industry, but who is willing to repeat erroneous statements about RDRAM on the internet. Most of your statements about RDRAM and SDRAM are too mushy for me to prove obviously erroneous. For example here's some more examples from back in February: "Rambus is cheaper for low end devices such as Timna based products because it uses less chips than eqivalent SDRAM." #reply-12853563 Okay, if Rambus is cheaper for Timna, how come Intel tried to bring Timna out with SDRAM instead of RDRAM, and then finally gave up on it completely? Mar 7, 2000 "Of the world's 7 major DRAM producers, 5 are in full production, Micron will be within a few weeks and Hitachi are in a legal dispute with Rambus." #reply-13096525 So who are these 5 major DRAM producers "in full production?" And when is Micron going to go into production again? Feb 29, 2000 Similarly we will soon see Samsung losing its Rambus head start and RDRAM prices coming down. #reply-12999392 If the real price of RDRAM had ever come down, Intel would have shipped Timna. If RDRAM prices were going to come down in 2001, Intel wouldn't be putting out an SDRAM solution in 3Q01. Feb 25, 2000 As supply increases along with yields and the number of manufacturers and fab lines, cost will come down. You are right about Samsung having a rare control over RDRAM price. They have almost a monopoly (but are not supplying PlayStation). This is a short term scenario and they will have to give in to market forces - forcing prices down as competitors ramp up. And about DDR RAM. It is not available. If it is so good, why aren't the major DRAM players ramping it? Surely they know more about yields, pricing and performance than we do. #reply-12973552 That was nearly a year ago. RDRAM prices are still ridiculously high, and now DDR is available from multiple retail vendors as the major DRAM players ramp it up. What part of this prediction is going to come true, and when? "Have you heard of the Timna chip for the sub $500 dollar PC (Celeron's successor)? It talks native RMBS. They are having to put a SDRAM converter on it because at the moment RMBS prices are too high because of demand! Intel knows what the state of the RMBS manufacturing ramps are - that is why Timna is a RMBS product. Did you see Samsung's announcement to build a sub $200 PC using exclusively Rambus?" #reply-12972412 Okay. If "Intel knows what the state of the RMBS [sic] manufacturing ramps are", then why did they can Timna, and why are they bothering to produce SDRAM support for the P4? And where is that sub $200 PC from Samsung, anyway? And from the same post: The whole DRAM and microprocessor industry knows that the future is Rambus. Even Sun has announced that Rambus is the memory to be used in its MAJC chip - another low end, cheap Rambus application #reply-12972412 The fact that you think that MAJC (which still hasn't appeared yet and is quite late) is a "low end, cheap" application indicates quite clearly that you weren't listening (or did not understand) when they gave out the production descriptions for the chip. MAJC is a very high end chip intended for relatively small production with very high prices. But not being connected to the industry, you're unaware of this. Here's a hint. Do you see Nvidia announcing any RDRAM compatible products? Would you imagine that there is a possibility that the reason is that Nvidia sells a low end cheap graphics card and can't afford the RDRAM premium? Yep, that's the ticket. Same reason Nintendo had to drop RDRAM for their next generation playstation. Same reason Microsoft never considered it. Feb 24, 2000 "Now Hyundai is ramping RMBS as fast as possible in addition to Toshiba, Samsung, INfineon etc. These DRAM giants must know something. I haven't seen much DDR RAM out of them yet. In fact DDR was rejected for the Playstation by SONY because it was too unstable. ... Intel last week called DDR "too little, too late" and "dead" as far as the workstation and small server market was concerned." #reply-12967559 Obviously the Hyundai ramp (as fast as possible), as well as the Infineon and Micron conversions to RDRAM are a bit delayed. As far as DDR being unstable, that didn't seem to stop Nvidia, Microsoft or Nintendo from designing the stuff in. Heck, if Intel knew that DDR was dead back in Feb 2000, why are they including it now in their plans for 2002? Does seem kind of odd, doesn't it. -- Carl