MIRGE: An Array-Based Computational Framework for Scientific Computing

TL;DR

MIRGE is a flexible, array-based computational framework that uses lazy evaluation and data-flow graphs to optimize scientific computing tasks across diverse hardware, enhancing performance and portability.

Contribution

The paper introduces MIRGE, a novel array computation framework that integrates lazy evaluation, domain-specific optimizations, and heterogeneous execution support for scientific computing.

Findings

MIRGE achieves high performance on GPU hardware.

The framework enables performance portability across different computing environments.

MIRGE facilitates scientific expressivity with efficient execution.

Abstract

MIRGE is a computational approach for scientific computing based on NumPy-like array computation, but using lazy evaluation to recast computation as data-flow graphs, where nodes represent immutable, multi-dimensional arrays. Evaluation of an array expression is deferred until its value is needed, at which point a pipeline is invoked that transforms high-level array expressions into lower-level intermediate representations (IR) and finally into executable code, through a multi-stage process. Domain-specific transformations, such as metadata-driven optimizations, GPU-parallelization strategies, and loop fusion techniques, improve performance and memory efficiency. MIRGE employs "array contexts" to abstract the interface between array expressions and heterogeneous execution environments (for example, lazy evaluation via OpenCL, or eager evaluation via NumPy or CuPy). The framework thus enables performance portability as well as separation of concerns between application logic, low-level implementation, and optimizations. By enabling scientific expressivity while facilitating performance tuning, MIRGE offers a robust, extensible platform for both computational research and scientific application development. This paper provides an overview of MIRGE. We further describe an application of MIRGE called MIRGE-Com, for supersonic combusting flows in a discontinuous Galerkin finite-element setting. We demonstrate its capabilities as a solver and highlight its performance characteristics on large-scale GPU hardware.