Sequential-modular-based process flowsheeting software remains an indispensable tool for process design, control, and optimization. Yet, as the process industry advances in intelligent operation and maintenance, conventional sequential-modular-based process-simulation techniques present challenges regarding computationally intensive calculations and significant central processing unit (CPU) time requirements, particularly in large-scale design and optimization tasks. To address these challenges, this paper proposes a novel process-simulation parallel computing framework (PSPCF). This framework achieves layered parallelism in recycling processes at the unit operation level. Notably, PSPCF introduces a groundbreaking concept of formulating simulation problems as task graphs and utilizes Taskflow, an advanced task graph computing system, for hierarchical parallel scheduling and the execution of unit operation tasks. PSPCF also integrates an advanced work-stealing scheme to automatically balance thread resources with the demanding workload of unit operation tasks. For evaluation, both a simpler parallel column process and a more complex cracked gas separation process were simulated on a flowsheeting platform using PSPCF. The framework demonstrates significant time savings, achieving over 60% reduction in processing time for the simpler process and a 35%–40% speed-up for the more complex separation process.