18265300. COMPILING DEVICE, COMPILING METHOD, AND COMPILING PROGRAM RECORDING MEDIUM simplified abstract (NEC Corporation)

From WikiPatents
Jump to navigation Jump to search

COMPILING DEVICE, COMPILING METHOD, AND COMPILING PROGRAM RECORDING MEDIUM

Organization Name

NEC Corporation

Inventor(s)

Kento Iwakawa of Tokyo (JP)

COMPILING DEVICE, COMPILING METHOD, AND COMPILING PROGRAM RECORDING MEDIUM - A simplified explanation of the abstract

This abstract first appeared for US patent application 18265300 titled 'COMPILING DEVICE, COMPILING METHOD, AND COMPILING PROGRAM RECORDING MEDIUM

Simplified Explanation

The patent application describes a compiling device, method, and program that can increase the speed of a program by optimizing loops that include access to adjacent data.

  • Loop analysis unit determines if vectorization of a loop in a program is possible when there is access to adjacent data.
  • Vectorization execution unit generates instructions to read elements required for computation from memory to a vector register when vectorization of the loop is possible.

---

      1. Potential Applications
  • High-performance computing
  • Data processing applications
  • Scientific simulations
      1. Problems Solved
  • Slow execution of programs with loops accessing adjacent data
  • Inefficient memory access patterns
  • Limited speed and performance of programs
      1. Benefits
  • Increased speed and efficiency of programs
  • Optimized memory access
  • Improved overall performance of applications


Original Abstract Submitted

Provided are a compiling device, a compiling method, and a compiling program with which the speed of a program can be increased when a loop in a program includes an access to adjacent data. The compiling device includes: a loop analysis unit for determining whether vectorization of a loop in a source program is possible when an access to adjacent data is included in the loop; and a vectorization execution unit for generating an instruction for reading elements required for computation from a memory to a vector register when access to the adjacent data is included and the vectorization of the loop is possible.