Intel 80386 Architecture

The Intel 80386 is a powerful 32-bit microprocessor and is the first hardware platform on which OS/2 Version 2.0 has been implemented. The 80386 incorporates multitasking support, sophisticated memory management, pipelined architecture, address translation caching, and a high-speed bus interface, all combined within the processor chip. While the 80386 represents a significant improvement over previous generations of Intel microprocessors, it retains software compatibility with older 16-bit microprocessors such as the 8086 and 80286 families.

The capacity of the 80386 processor is significant. Some figures are presented below, in comparison with 80286 processors:

  • or 6 million instructions per second (corresponding to clock speeds of 16, 20, and 25 MHz).

  • gigabyte physical address space, compared with the maximum of 16 megabytes available on the 80286.

    Note: The 80386SX is an exception as it still is limited to a maximum of 16MB of physical memory. Internally the 80386SX is a full 32-bit processor but externally it has only a 16-bit data bus and a 24-bit address bus.

  • 64 terabyte virtual address space, compared with the 1 gigabyte available on the 80286.

  • Ability to handle memory objects from 1 byte to 4 gigabytes in size, compared to segments of 16 bytes to 64 kilobytes on the 80286.

  • Paged memory management using 4 kilobyte pages, compared to the 80286 which offered only segmented memory management.

    This chapter provides an overview of the 80386 processor architecture, in order to serve as a base for understanding the changes made in OS/2 Version 2.0. More detailed information about the 80386 can be found in:

  • Intel 80386 Hardware Reference Manual (ISBN 1-55512-069-5)

  • Intel 80386SX Hardware Reference Manual (ISBN 1-55512-105-5)

  • Intel 386 DX Programmer's Guide (ISBN 1-55512-082-2)

  • IBM PS/2 Model 80 Technical Reference (IBM P/N 84X1508).

    [Back: Summary]
    [Next: Physical Characteristics]