How it works
A microprocessor (e.g., Intel 8086, ARM Cortex-A53) contains only the CPU — ALU, registers, control unit — and requires external memory and peripheral chips. It has a wide external bus (16–64 bit data bus), operates at high clock speeds (1 MHz to several GHz), and is optimised for computing throughput. A microcontroller (e.g., 8051, ATmega328P, PIC16F877A, STM32F103) integrates CPU, program memory (flash), data memory (RAM/EEPROM), I/O ports, timers, ADC, communication interfaces on one chip. The 8051 has 4 KB ROM, 128 bytes RAM, 4 × 8-bit I/O ports, 2 timers, and a UART on-chip. ATmega328P runs at up to 20 MHz at 3.3–5 V, consuming under 1 mA in active mode and under 1 μA in power-down sleep mode.
Key points to remember
The key architectural distinction: microprocessor is brain-only, microcontroller is brain-plus-body. Microcontrollers are designed for real-time, low-power, embedded control; microprocessors are designed for maximum computation throughput in systems with abundant memory and peripherals. Clock speed alone does not distinguish them — a modern ARM Cortex-M4 microcontroller at 168 MHz outperforms a 1990s Pentium microprocessor in raw MIPS. Von Neumann microcontrollers (8051) use a single bus for program and data memory; Harvard microcontrollers (AVR ATmega, PIC) use separate buses, allowing simultaneous instruction fetch and data access for higher throughput. On-chip ADC resolution is typically 10–12 bits for microcontrollers; external ADC chips are needed for 16-bit or higher precision.
Exam tip
The examiner always asks you to compare microcontroller and microprocessor in a table with at least five parameters — structure your table with rows: memory, peripherals, cost, power, application, and bus width; fill all entries; losing one row means losing one mark.