2. The 4004 needed EROM for instruction memory and DRAM for data memory. If those were not available, there would not have been a microcomputer chip set. In other words, semiconductor memory played a vital role in complementing the CPU on a chip- without which there would not have been a microprocessor market in the early to mid 1970s!
3. Marcian (Ted) Hoff Jr was the Intel systems engineer working with all the semiconductor process guys to make the designs work in practice as well as writing about semiconductor memory applications in various types of systems. Without that expertise, and teamwork, Intel would not have been as successful as they were in semiconductor memory (Ted's story about the 1102 DRAM- flawed pre-cursor to the 1103- illustrates that point- the process guys needed systems engineer to get it right).
Here are the 6 sub-topics we will discuss during the fireside chat:
1. Describe the environment at Intel from Sept 1968 till late 1971 as it relates to the leaders (Noyce, Moore, etc), the bold opinion that semiconductor memories would replace core memories (at a cost of 1 cent/bit), the advancement of the MOS and bipolar semiconductor processes, etc. Why was it "the greatest learning experience of a lifetime?"
2. Who were Intel's semiconductor memory competitors in the early 1980s and what gave Intel a competitive edge/advantage?
3. What types of semiconductor memory were used to complement the 4004 and 8008, i.e. instruction memory, data memory, pushdown stack, etc.
4. What did Ted do to demonstrate how semiconductor memories could be used for :
-core memory replacement for computers
-register files for industrial applications
-embedded controller applications that used microprocessors instead of random logic
5. What was the significance of the Silicon Gate process at Intel in terms of semiconductor reliability, performance and manufacturability (decent yield per semiconductor wafer)? How did Intel effectively use that process to make p and n channel MOS as well as bipolar LSIs?
6. What is the current status and future for solid state memories? Are we close to the limits of Moore's law in terms of density, power consumption, reliability, etc for Flash and newer semiconductor memory technologies?