A Domain-Specific Language and Editor for Parallel Particle Methods

Abstract

Domain-specific languages (DSLs) are of increasing importance in scientific high-performance computing to reduce development costs, raise the level of abstraction, and, thus, ease scientific programming. However, designing DSLs is not easy, as it requires knowledge of the application domain and experience in language engineering and compilers. Consequently, many DSLs follow a weak approach using macros or text generators, which lack many of the features that make a DSL comfortable for programmers. Some of these features—e.g., syntax highlighting, type inference, error reporting—are easily provided by language workbenches, which combine language engineering techniques and tools in a common ecosystem. In this article, we present the Parallel Particle-Mesh Environment (PPME), a DSL and development environment for numerical simulations based on particle methods and hybrid particle-mesh methods. PPME uses the Meta Programming System, a projectional language workbench. PPME is the successor of the Parallel Particle-Mesh Language, a Fortran-based DSL that uses conventional implementation strategies. We analyze and compare both languages and demonstrate how the programmer’s experience is improved using static analyses and projectional editing, i.e., code-structure editing, constrained by syntax, as opposed to free-text editing. We present an explicit domain model for particle abstractions and the first formal type system for particle methods.

A Mathematical Definition of Particle Methods

“…Future work could also attempt a less monolithic definition that allows modular combinations of different particle methods. While this could lead to a formulation that can potentially directly be exploited in software engineering or in the design of domain-specific programming languages for particle methods [22], it does require solving a few theoretical problems: One such problem is how to synchronize data when different particle methods access shared particles. Other open issues include how different types of particles from different methods can interact with one another (e.g., for remeshing) and how access of an algorithm can be restricted to a subset of particles (e.g., for boundary element methods).…”

“…Also simulations of hybrid discrete-continuous models are possible, as often done in plasma physics, where discrete point charges are coupled with continuous electric and magnetic fields [18]. In addition to their versatility, particle methods can also efficiently be parallelized on shared-and distributed-memory computers [22,20,32,21,31], and they simplify simulations in complex [33] and time-varying [3] geometries, as no computational mesh needs to be generated and maintained. Particle methods have therefore been widely used in numerical simulations.…”

Easy and efficient agent-based simulations with the OpenABL language and compiler

“…Vasista et al [63] introduces a DSL based on the PolyMage framework for geometric multigrid methods. Similarly, Karol et al [64] shows a DSL for parallel particle methods.…”

EVEREST: A design environment for extreme-scale big data analytics on heterogeneous platforms

“…Application experts are offered embedded domain-specific languages (DSLs) to express the semantics and security requirements of computational tasks to enable high-level code optimizations. DSL extensions have been successfully demonstrated in many domains, such as computational fluid dynamics [12], hybrid particle-mesh simulations [13], tensor expression optimizations [14]- [16], and dataflow languages [17]. EVEREST proposes a data-centric approach for security, dealing with confidentiality, authentication and integrity of the data handled by the system with hardware-assisted data protection applied to both edge devices and data center nodes.…”