Intel 80C186EA, User Manual

Page: 298 / 417

11-1
CHAPTER 11
MATH COPROCESSING
The 80C186 Modular Core Family meets the need for a general-purpose embedded microproces-
sor. In most data control applications, efficient data movement and control instructions are fore-
most and arithmetic performed on the data is simple. However, some applications do require
more powerful arithmetic instructions and more complex data types than those provided by the
80C186 Modular Core.
11.1 OVERVIEW OF MATH COPROCESSING
Applications needing advanced mathematics capabilities have the following characteristics.
•
Numeric data values are non-integral or vary over a wide range
•
Algorithms produce very large or very small intermediate results
•
Computations must be precise (i.e., calculations must retain several significant digits)
•
Computations must be reliable without dependence on programmed algorithms
•
Overall math performance exceeds that afforded by a general-purpose processor and
software alone
For the 80C186 Modular Core family, the 80C187 math coprocessor satisfies the need for pow-
erful mathematics. The 80C187 can increase the math performance of the microprocessor system
by 50 to 100 times.
11.2 AVAILABILITY OF MATH COPROCESSING
The 80C186 Modular Core supports the 80C187 with a hardware interface under microcode con-
trol. However, not all proliferations support the 80C187. Some package types have insufficient
leads to support the required external handshaking requirements. The 3-volt versions of the pro-
cessor do not specify math coprocessing because the 80C187 has only a 5-volt rating. Please refer
to the current data sheets for details.
To execute numerics instructions, the 80C186EA must exit reset in Numerics Mode. The proces-
sor checks its TEST pin at reset and automatically enters Numerics Mode if the math coprocessor
is present.