ParEval

The 'ParEval' dataset/benchmark contains a curated collection of tasks designed to evaluate the performance and correctness of synthesized parallel algorithms for irregular data structures.

Papers using ParEval (3)

ParEVO: Synthesizing Code for Irregular Data: High-Performance Parallelism through Agentic Evolution2026ParaCodex: A Profiling-Guided Autonomous Coding Agent for Reliable Parallel Code Generation and Translation2026OMPar: Automatic Parallelization with AI-Driven Source-to-Source Compilation2024 · 2 cites