Heterogeneous Parallel Primitives Programming Model

Abstract

With the success of programming models such as OpenCL and CUDA, heterogeneous computing platforms are becoming mainstream. However, these heterogeneous systems are low-level, not composable, and their behavior is often implementation defined even for standardized programming models. In contrast, the method and system embodiments for the heterogeneous parallel primitives (HPP) programming model disclosed herein provide a flexible and composable programming platform that guarantees behavior even in the case of developing high-performance code.

Description

RELATED APPLICATIONS[0001]This application is related to the U.S. Provisional Patent Application No. 61 / 652,772, filed on May 29, 2012, which is incorporated by reference herein in its entirety.BACKGROUND[0002]1. Field of the Invention[0003]The present invention relates generally to a programming model for a heterogeneous processor system.[0004]2. Background Art[0005]With the success of programming models such as OpenCL and CUDA, heterogeneous computing platforms are becoming mainstream. However, these heterogeneous systems are low-level, not composable, and their behavior is often implementation defined even for standardized programming models.[0006]Thus what is needed are system and method for a heterogeneous parallel primitives (HPP) programming model that provides a flexible and composable programming platform that guarantees behavior even in the case of developing high-performance code.SUMMARY OF EMBODIMENTS[0007]According to an embodiment, a method and system for executing an ...

AI Technical Summary

Benefits of technology

The patent describes a method and system for executing asynchronous tasks on a heterogeneous computing platform. An asynchronous task is initialized and results that become available during execution are encapsulated. The task is executed on a grid and the result is assigned to it when it becomes available. The patent also describes a system for managing memory, where a distributed array is generated and bound to a kernel that executes a workgroup on a processor in a heterogeneous computing platform. The bound distributed array is accessed by the workgroup during execution. Overall, this invention allows for efficient and effective execution of tasks on different types of processors and memories, improving overall performance and efficiency of heterogeneous computing platforms.

Problems solved by technology

However, these heterogeneous systems are low-level, not composable, and their behavior is often implementation defined even for standardized programming models.

Images