Motorola 68882 (MC68882) is an enhanced version of
68881 Floating Point Unit. The 68882 is
pin- and object-code compatible with the 68881, and, like the 68881,
it could work with Motorola 68020 and
68030 microprocessors.
The architecture of the 68882 FPU is similar to the 68881. The
FPU has 8 directly-addressable 80-bit registers. It supports seven
data types: 8-bit, 16-bit and 32-bit integers, single, double and
extended-precision floating point numbers, and packed decimal
strings. The 68882 always converts external operands into internal
extended-precision format, and performs all calculations in
this format. The MC68882 has the same instruction set as the 68881
co-processor.
The 68882 FPU includes a few features that significantly improve
its performance:
- The 68882 may execute more than one floating-point instruction at
the same time.
- The MC68882 uses faster conversion method of external operands
to/from FPU's internal format.
To get maximum performance from the co-processor the programs
needs to be re-compiled specifically for the 68882. The 68882 may run
re-compiled programs up to two times faster than the 68881 FPU.
Performance gain for older applications, i.e. applications compiled
for Motorola 68881 co-processor, is usually is not as impressive -
the FPU may run these programs up to 15 - 25% faster than the 68881.
|
