Carl, I am not in the memory business and as such, I am lucky that "deliberations" are carried out in absolute ignorance of the facts (g). Yet, despite the paucity of my understanding of the intricacies, I doubt that a conservative (engineering wise) company like INTC would have gone with DRAM revolution rather than evolution. After all, in the years that they were cooperating, the DDR solution which was patented as well (the douyble clocking) effectively would have doubled the bandwidth, INTC must have thought that they will need much more than just doubling and that RMBS can provide this at a Si real estate penalty of about 15%. Additional problems related to board designs and bus designs that require very tight specifications, are really not a road block, it could be a bump, and whenever you got to extreme frequencies you got to pay attention to that (many years back, in the early 60' when I was doing my Matser thesis, I had to design one of the first nanosecond coincidence circuits, I must have spent 25% of my time debugging that poor circuit to death). The DDR MOBO will have similar challenges, IMHO.
Zeev |
