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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Paul Engel who wrote (82137)	5/30/1999 4:26:00 PM
From: Scumbria	Read Replies (3) \| Respond to of 186894

Paul, Here is what a noted authority said about DRDRAM earlier this year. Too bad Intel didn't listen to him. ;^(( The reality is that no CPU needs DRDRAM. CPU's burst data in and out 32 bytes at a time. That is not enough data to get any benefit out of high bandwidth memory. The only real advantage of DRDRAM over SDRAM is a smaller pin count. High bandwidth is needed for graphics, not CPU's. The whole Intel/Rambus connection is little more than a marketing ploy. Message 7442772 Message 7744162 techstocks.com Message 7483131 techstocks.com Scumbria

To: Paul Engel who wrote (82137)	5/30/1999 5:23:00 PM
From: kash johal	Read Replies (3) \| Respond to of 186894

Paul, Re: "At this point, the RAMBUS jeopardy seems to be focusing not on functionality - Intel seems to have resolved those issues - but on performance. If RAMBUS provides minimal, or no, performance advantage over PC133 SDRAMS, then the issue quickly becomes one of COST ! The Coppermine/Camino/RAMBUS solution then goes UP in price since the RAMBUS memory will cost more than SDRAM memory. How, then, will Intel deal with this COST ADDER ?" I think there is a motherboard jeopardy as well. Intel went to the slot archirecture beacuase L2 cache rams needed to be close to the CPU. So instead of having them on the PC board they were embedded onto the CPU package for quality and reliability. I think the MB costs for Rambus will be much higher due to the manufacturing problems of handling the RIMMS away from the CPU. Perhaps they should have gone to a slot design or something for coppermine. They would have the coppermine and the RDRAM on the same module. They could sell several options from 128Mb RAM to 512Mb RAM. Having 800Mhz signal wandering all over the board creates major problems in design/manufacturing and quality. I am sure the problems will be solved but it may well take 3-6 months. However a 3 month delay would cause them to miss the all important Q4 selling season. In the worst case I imagine they can still sell Coppermine with 100Mhz SDRAM - I think. Several folks have documented that the RDRAM doesn't buy much over SDRAM anyway- so they still should have pretty good performance with regular SRDRAM and Coppermine. Obviously it would be a disaster for Rambus if this blows up, but other than it bruising the egoes at folks at Intel it may not be a big deal at all. Regards, Kash.

To: Paul Engel who wrote (82137)	5/30/1999 6:57:00 PM
From: grok	Read Replies (1) \| Respond to of 186894

Re: <There used to be talk of an interface chip that would allow SDRAM modules to be plugged in to RIMM sockets - sitting between the Camino Chip Set and the RIMM sockets, thereby allowing SDRAM memory to be used with the Camino. I sure hope that "solution" is available, but I haven't heard much about this in recent months.> It is appropriate for you to put "solution" in quotes because it is an awful solution since it adds cost and latency over a straight sdram solution. Paul, do you think that it is possible that Intel has a secret 44xxx chip set for Coppermine in the back room with direct interface to sdram that they been keeping quiet so as to not take the pressure off the dram industry to switch to drdram? Could they do such a thing and keep it secret?