SI
SI
discoversearch

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: jim kelley who wrote (36869)2/3/2000 4:01:00 PM
From: Tenchusatsu  Read Replies (1) | Respond to of 93625
 
Jim, <Sounds like the server attitude is driven by RDRAM costs.>

It's more than that, although the cost is the biggest reason. SDRAM allows for double-sided, double-stacked modules which cram 32 devices onto one DIMM. And with up to four DIMMs per SDRAM channel, the maximum capacity per channel is 128 devices.

RDRAM, on the other hand, has a protocol limit of 32 devices per channel. To have more devices, you have to use hubs and turn a single main RDRAM channel into several branch channels. But I hear that solution isn't very desirable to server guys.

<I suspect that the initial DDR and double DDR offering will also carry a premium cost because they are new products and do not yet have economies of scale.>

That may be true. Also, notice how DRAM companies like Micron are releasing both 200 and 266 MHz versions of DDR. DDR-200 is NOT going to be competitive performance-wise against RDRAM at all. DDR needs to run at 266 in order to present a viable performance alternative to RDRAM, but it seems the mere existence of DDR-200 means DDR-266 isn't as easy to manufacture.

On the other hand, it's generally accepted that DDR will indeed be cheaper than RDRAM, even if the difference in the long run is on the order of 10% or so.

Tenchusatsu