eeNews: You were in-charge of memory at one time. Do you ever think SGS-Thomson…or ST should have stayed in memory?
Carlo Bozotti: I was in charge of memory for about six years. We did try to get bigger. We were already doing EPROM, EEPROM, NOR-flash. At the end of the 1990s we looked to expand into NAND. We tried hard to bring in memories but the NOR-flash business was not sustainable alone. The objective was to become bigger in all forms of memory except for computer memory, so that was NAND and low-power DRAM at the end of the 1990s.
In the end, we entered the Numonyx joint-venture with Intel as a deconsolidation strategy, and then we sold to Micron. I believe this was the right move. In parallel, we managed the conversion of our two big memory fabs that are now making BCD.
eeNews: Was the problem the capital intensity required to stay in memory?
Carlo Bozotti: It was partly the capital intensity but we did not have the real technical advantage in low-power DRAM and NAND.
Carlo Bozotti: Europe must continue to lead.
eeNews: Are there any decisions that you made, or that the company made, that make you stop and think how things could have been different?
Carlo Bozotti: When I started as CEO we wanted to be an important player in two domains. One is in digital platforms and the second is supporting components such as BCD, analog, MEMS, sensors and power.
In microcontroller platforms – our STM32 family – we have done well and the ecosystem is very broad. We wanted to do the same in application processors for mobile and consumer. If there is one thing I regret it is not taking action on [to exit] set-top box (STB) sooner. I should have acted quicker. This I would change.