Intel 80486 microprocessor family
The successor to the 80386 processor
Intel 80486 (i486) included many changes to its microarchitecture
that resulted in significant performance improvements:
- 8 KB unified level 1 cache for code and data was added to the CPU.
In later versions of the 80486 the size of level 1 cache was
increased to 16 KB.
- Execution time of instructions was significantly reduced. Many
load, store and arithmetic instructions executed in just one cycle
(assuming that the data was already in the cache).
- Intel 486 featured much faster bus transfers - 1 CPU cycle as
opposed to two or more CPU cycles for the 80386 bus.
- Floating-point unit was integrated into 80486DX CPUs. This
eliminated delay in communications between the CPU and FPU.
Furthermore, all floating-point instructions were optimized - they required
fewer number of CPU cycles to execute.
- Clock-doubling and clock-tripling technology was introduced in
faster versions of Intel 80486 CPU. These i486 processors could run
in existing motherboards with 20 - 33 MHz bus frequency, while running
internally at two or three times of bus frequency. 80486SX2 and
80486DX2 were clock-doubled version, and 80486DX4 was a clock-tripled
version. AMD also produced 80486DX5 or X5 - clock-quadrupled version of the
- Power management features and System Management Mode (SMM) became a standard
feature of the processor.
A few different variations of the 80486 microprocessors were
produced. Two most common versions are 80486DX with integrated FPU
and 80486SX without integrated FPU. There were also low power
versions and embedded 80486 microprocessors.
Intel 80486 microprocessor was produced at speeds up to 100 MHz. AMD
produced even faster 120 and 133 MHz versions of the 80486, and
manufactured in small quantities 150 MHz and possibly 166 MHz
Picture of: Cyrix Cx486DX-V33QP
Picture of: IBM 486-4V375QIC
Picture of: Texas Instruments TI486SXL2-50
168-pin ceramic PGA
Picture of: UMC U5SX 486-33F /40
33 (40?) MHz