Carver Mead and Lynn Conway published Introduction to VLSI Systems ca. 1980, and MOSIS started in 1981. BBN was interested in developing design tools and understanding the potential market for these services, so allocated some funds for some internal development projects.
We were advised to do the clock generation off the chip because "everybody screws it up the first time." I didn't, and I did, so the prototype chips only partially worked. Unfortunately, we never got a chance at a second round to fix the problems as BBN provided no more funding.
MOSIS is still running, though today it appears to be very much CMOS oriented (the chip described here used NMOS technology).