Furthermore, in order for the students to experience the complete board-level software design cycle, they write 8085 code to perform a certain task, assemble it, and burn the resulting hexadecimal code into an EPROM. This memory chip is then added to the 8085-based microcomputer board.
Both an I/O chip for serial communication with the development system and an EPROM would be incorporated.
Students would be encouraged to develop their controllers on the PC, then download their code onto the board-level microprocessor via an EPROM. Other interesting projects might involve microprocessor-based signal processing or developing a warning system based on several different input conditions.
During this stage of the research, Intel top management also made important decisions regarding EPROMs and microprocessors.
For EPROMs and microprocessors, historical data were combined with current data obtained during the research period.
Cross-case analysis suggested that the stages were roughly the same for DRAMs and EPROMs but that the time taken by each stage was generally longer for EPROMs.
The fourth question--If not planned by top management, how did microprocessors and EPROMs come about at Intel in the first place?--focused attention on the evolution of Intel's distinctive competence, which produced unanticipated innovations, and on the role played by top management in the internal selection processes in supporting these innovations.
The fifth question--If Intel exited from DRAMs because they had become a commodity business, how would that affect the future of EPROMs, which had also become a commodity business by 1988?--was answered during the second stage of the research (1990-1991): Intel decided to halt expansion of manufacturing capacity in EPROMs in 1991.
EPROMs and microprocessors were unplanned new technologies with major commercial potential.
For instance, while DRAMs were estimated to account for only about 3 percent of Intel's $1.6 billion sales revenues in 1984, budgeted expenditures for DRAM TD for 1985 were estimated to be roughly $65 million--one third of total R&D and about the same as for EPROMs and microprocessors each (Cogan and Burgelman, 1990).
Radiation effects on EPROM and EEPROM semiconductor memories have been tested by exposing samples to gamma radiation, the source of radiation being [sup.60]Co.
Figures 1 and 2 show the average relative change in number of errors versus the absorbed dose of radiation, in irradiated EPROM and EEPROM samples, respectively.