Short answers, pulled from the story.
An application-specific integrated circuit is a customized chip built to perform a single task rather than many. Unlike general-purpose chips found in digital voice recorders or video codecs, these devices are designed for particular uses from the start.
Ferranti and Interdesign were manufacturing early bipolar gate arrays by 1967. This event marked the beginning of specialized chip history alongside Fairchild Semiconductor introducing the Micromatrix family that same year.
The maximum complexity possible has grown from 5,000 logic gates to over 100 million as feature sizes shrank and design tools improved. Modern versions often include entire microprocessors alongside memory blocks like ROM, RAM, EEPROM, and flash memory.
Structured ASIC design reduces both manufacturing cycle time and design cycle time compared to cell-based approaches by virtue of pre-defined metal layers. These structured devices often integrate large blocks of system functionality including digital signal processor units and peripherals rather than only functional units.
Complementary metal oxide semiconductor technology opened the door to broad commercialization when Robert Lipp developed the first CMOS gate arrays in 1974 for International Microcircuits Inc. Metal oxide semiconductor standard-cell technology followed under trade names Micromosaic and Polycell during the 1970s.