18463142. METHODS AND APPARATUS TO INSERT PROFILING INSTRUCTIONS INTO A GRAPHICS PROCESSING UNIT KERNEL simplified abstract (Intel Corporation)

From WikiPatents
Jump to navigation Jump to search

METHODS AND APPARATUS TO INSERT PROFILING INSTRUCTIONS INTO A GRAPHICS PROCESSING UNIT KERNEL

Organization Name

Intel Corporation

Inventor(s)

Konstantin Levit-gurevich of Kiryat Byalik (IL)

Orr Goldman of Harel (IL)

METHODS AND APPARATUS TO INSERT PROFILING INSTRUCTIONS INTO A GRAPHICS PROCESSING UNIT KERNEL - A simplified explanation of the abstract

This abstract first appeared for US patent application 18463142 titled 'METHODS AND APPARATUS TO INSERT PROFILING INSTRUCTIONS INTO A GRAPHICS PROCESSING UNIT KERNEL

Simplified Explanation

The patent application describes a method for inserting profiling instructions into graphics processing unit (GPU) kernels. The method involves determining if a GPU supports modification of entry point addresses and detecting the original entry point addresses of a GPU kernel. Based on this determination, an instrumented GPU kernel is created by inserting profiling initialization instructions and jump instructions at the entry point addresses, as well as inserting profiling measurement instructions into the instrumented GPU kernel.

  • The method determines if a GPU supports modification of entry point addresses.
  • The original entry point addresses of a GPU kernel are detected.
  • Profiling initialization instructions and jump instructions are inserted at the entry point addresses of the instrumented GPU kernel.
  • Profiling measurement instructions are inserted into the instrumented GPU kernel.

Potential Applications

  • Performance analysis and optimization of GPU kernels.
  • Debugging and profiling of GPU applications.
  • Monitoring and analyzing GPU performance in real-time.

Problems Solved

  • Lack of efficient profiling and debugging tools for GPU kernels.
  • Difficulty in analyzing and optimizing GPU performance.
  • Limited visibility into the execution of GPU applications.

Benefits

  • Improved performance analysis and optimization of GPU kernels.
  • Enhanced debugging and profiling capabilities for GPU applications.
  • Real-time monitoring and analysis of GPU performance.


Original Abstract Submitted

Embodiments are disclosed for inserting profiling instructions into graphics processing unit (GPU) kernels. An example apparatus includes instructions, and at least one processor to execute the instructions to determine whether a GPU supports modification of entry point addresses, detect a first entry point address and a second entry point address of an original GPU kernel, create a corresponding instrumented GPU kernel from the original GPU kernel based on the determination by inserting at least one of first profiling initialization instructions or first jump instructions at the first entry point address of the instrumented GPU kernel, inserting at least one of second profiling initialization instructions or second jump instructions at the second entry point address of the instrumented GPU kernel, and inserting profiling measurement instructions into the instrumented GPU kernel.