IntervalMDP.jl: Accelerated Value Iteration for Interval Markov Decision Processes

TL;DR

IntervalMDP.jl is a Julia package that accelerates the analysis of interval Markov Decision Processes using multi-threaded and GPU-accelerated value iteration, outperforming existing tools in efficiency.

Contribution

The paper introduces a new Julia package that significantly improves the speed and memory efficiency of IMDP analysis through parallelized value iteration algorithms.

Findings

Substantially faster verification and synthesis of strategies.

Reduced memory consumption compared to existing tools.

Effective utilization of GPU and multi-core CPU for IMDP analysis.

Abstract

In this paper, we present IntervalMDP.jl, a Julia package for probabilistic analysis of interval Markov Decision Processes (IMDPs). IntervalMDP.jl facilitates the synthesis of optimal strategies and verification of IMDPs against reachability specifications and discounted reward properties. The library supports sparse matrices and is compatible with data formats from common tools for the analysis of probabilistic models, such as PRISM. A key feature of IntervalMDP.jl is that it presents both a multi-threaded CPU and a GPU-accelerated implementation of value iteration algorithms for IMDPs. In particular, IntervalMDP.jl takes advantage of the Julia type system and the inherently parallelizable nature of value iteration to improve the efficiency of performing analysis of IMDPs. On a set of examples, we show that IntervalMDP.jl substantially outperforms existing tools for verification and strategy synthesis for IMDPs in both computation time and memory consumption.