Question

Describe the reasons why techniques for parallel execution of integer operations do not always work well for floating point arithmetic in MIPS

Answer 1

We know that in MIPS we can perform integer operations and floating-point operations. But in case of execution of Integer operation parallelly we can perform it but in the case of floating-point arithmetic it always not work well because as we know that in floating-point arithmetic we have to do an extra operation like rounding off, single precision, double precision, etc . For that different operation in floating-point arithmetic takes different times to execute, as a result, it becomes inefficient and it always not hold good . But in the case of integer operation, there is no extra overhead of like that .So they can perform parallelly efficiently.