Intel® Fortran Compiler 16.0 User and Reference Guide
High-level Optimizations (HLO) exploit the properties of source code constructs (for example, loops and arrays) in applications developed in high-level programming languages. While the default optimization level, option O2 , performs some high-level optimizations, specifying the O3 option provides the best chance for performing loop transformations to optimize memory accesses.
Loop optimizations may result in calls to library routines that can result in additional performance gain on Intel® microprocessors than on non-Intel microprocessors. The optimizations performed can also be affected by certain options, such as /arch (Windows*), -m (Linux* and OS X*), or [Q]x options. Additional HLO transformations may be performed for Intel® microprocessors than for non-Intel microprocessors.
Within HLO, loop transformation techniques include:
Loop Permutation or Interchange
Loop Distribution
Loop Fusion
Loop Unrolling
Data Prefetching
Scalar Replacement
Unroll and Jam
Loop Blocking or Tiling
Partial-Sum Optimization
Predicate Optimization
Loop Reversal
Profile-Guided Loop Unrolling
Loop Peeling
Data Transformation: Memset Combining, Memory Layout Change
Loop Rerolling
Memset and Memcpy Recognition
Statement Sinking for Creating Perfect Loopnests
Multiversioning: Checks include Pointer Dependency and Trip Counts
Loop Collapsing