To: jim kelley who wrote (44018 ) 6/10/2000 10:58:00 PM From: Ali Chen Read Replies (1) | Respond to of 93625
jim-kelley, you apparently are not reading well. <RDRAM is already in volume manufacturing. You talk as though it can't be made.> You are not listening. I said it requires outstanding efforts to keep RAMBUS within outrageous tolerances specified to "copy exactly" that lucky laboratory prototype. This drives the cost up and always will. <..then you assert with no evidence to support you that the DDR and DDR-II busses will be vastly superior.> First, the whole my message was a guide in support of the idea that to chain 30+ bi-directional devices on a single wire is a bad idea. Second, where did you find any DDR word in my post? Then, why do you think that DDR people will chain 30 DIMMs? <Go ahead and remove the pins and solder connections from your DDR chips if you can.> You still do not get it. You keep forgetting that DDR frequency is presumably 3 times lower than RMAbus. Three times lower frequency, eight times less devices, 20 times less problems at least. <Simply put, my point is the criticisms made by BILOW of the RDRAM transmission line bus are also true also of the DDR-II bus.> Simply put, I see that you got no clue as yet. No, your statement is not true; see the above. <I have done lots of engineering for high volume PC board manufacturing in my past lives. Never got stopped by any of these problems which you seem to think are almost insurmountable.> You seem to be one of those pointy-hair managers who thinks that if ten years ago they were not stopped by their problems, they will also defy the laws of physics even if they have no foggiest clue about the real cause of the problem, as you just have demonstrated: <The 820 motherboard according to INTEL specs is 4 layers. This is not a big deal.>