8087 Math Co-Processor ~ nerthinksjareen

A co-processor is a specially designed circuit on a microprocessor chip that can perform the same task very quickly that the microprocessor performs.
The main advantage of a co-processor is it reduces the workload of the main processor.
Co-processors share the same memory, I/O system, bus control, and clock generator.
The co-processor handles specialized tasks like mathematical calculation, graphical display on the screen, etc.

It was the first math co-processor designed by Intel to pair with 8086/8088 resulting in the easier and faster calculation.
The purpose of 8087 was to speed up the computation involving floating point calculation.
8087 math co-processor is also known as Numeric Data Processor (NDP), Numeric Processor Extension (NPX), Floating Point Unit (FPU)

This increase the overall speed and system performance of the entire system.
Addition, subtraction, multiplication, and division of simple numbers are not the coprocessor's job.
It does all the calculations involving floating point numbers like scientific calculation and algebraic functions.
Having a co-processor, which performs all the calculations, can free up a lot of CPU time.
This would allow the CPU to focus all of its resources on the other functions it has to perform.
It follows IEEE floating point standards.
This co-processor introduced about 60 new instructions available to the programmers.
All the mnemonics begin with 'F' to differentiate them from the standard 8086 instructions. e.g.FADD, FMUL.
If an instruction is an ESC-type instruction then the co-processor executes it else the microprocessor executes it.
8087 can work on integers, decimals, real, and floating point numbers.
It is capable of doing complex arithmetic and trigonometric calculations.
It is basically a secondary processor it has its own architecture and instruction set.

Floating point operations are complex math operations that required large registers, complex circuits, and an area on the chip.
A general processor avoids this much burden and delays such operation to a co-processor designed specifically for this purpose e.g. 8087 math co-processor for 8086 microprocessor.

Next: Architecture of 8087

nerthinksjareen