HiCR, an Abstract Model for Distributed Heterogeneous Programming

2025-09-01

AI Summary

To address the challenge of uniformly modeling hardware topology, kernel execution, memory management, communication, and instance management semantics across distributed heterogeneous systems, this paper proposes HiCR—a semantic abstraction model that introduces a Runtime Support Layer (RSL) positioned between hardware and runtime systems. HiCR employs a minimal, orthogonal set of abstract operations to capture essential semantics, decouples hardware specifics via a plugin-based architecture, and supports dynamic device binding and cross-platform execution. Its hardware-agnostic interface is compatible with mainstream parallel programming paradigms—enabling seamless integration with existing and future heterogeneous systems without code refactoring. Experimental evaluation across diverse heterogeneous platforms demonstrates HiCR’s strong portability and scalability, significantly reducing system migration and maintenance overhead.

Technology Category

• Machine Learning: Hardware-aware ML
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Data Mining & Knowledge Management: Scalability, Parallel & Distributed Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

We present HiCR, a model to represent the semantics of distributed heterogeneous applications and runtime systems. The model describes a minimal set of abstract operations to enable hardware topology discovery, kernel execution, memory management, communication, and instance management, without prescribing any implementation decisions. The goal of the model is to enable execution in current and future systems without the need for significant refactoring, while also being able to serve any governing parallel programming paradigm. In terms of software abstraction, HiCR is naturally located between distributed heterogeneous systems and runtime systems. We coin the phrase emph{Runtime Support Layer} for this level of abstraction. We explain how the model's components and operations are realized by a plugin-based approach that takes care of device-specific implementation details, and present examples of HiCR-based applications that operate equally on a diversity of platforms.

Problem

• Modeling distributed heterogeneous application semantics and runtime systems
• Enabling execution across current and future systems without refactoring
• Supporting diverse hardware through abstract operations and plugin approach

Innovation

• Abstract model for distributed heterogeneous programming
• Plugin-based approach for device-specific implementation
• Minimal operations enabling hardware discovery and execution