发表机构
Universitat Oberta de Catalunya (UOC); Barcelona Supercomputing Center (BSC)（开放大学加泰罗尼亚分校（UOC）; 巴塞罗那超级计算中心（BSC））
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本综述梳理了LLM在HPC编程五大类任务中的应用，指出通用LLM在串行等任务表现尚可但分布式范式不足，领域专用模型准确率更高但范围狭窄，LLM短期内无法取代HPC专家，将成为软件开发的强大合作者，二者融合是长期共同演化过程。
AI 中文摘要
大语言模型（LLM）正成为高性能计算（HPC）领域颇具潜力的助手，而HPC编程仍十分复杂且对专业知识要求极高。本综述系统梳理了LLM在五大类任务中的应用：代码生成、并行化与优化、框架与架构、评估与基准测试，以及更广泛的挑战。分析凸显了机遇与局限：通用型LLM在串行任务及类OpenMP任务上表现尚可，但在MPI等分布式范式中存在不足，而这类范式对正确性和可扩展性要求极高。领域专用模型（如HPC-Coder、HPC-GPT、chatHPC）通过微调、精心构建的数据集及检索增强生成（RAG）实现了更高的准确率，但其适用范围仍较狭窄，且评估大多局限于基准测试或微内核。整体来看，LLM兼具潜力与脆弱性：它们可降低入门门槛、加速原型开发并支持代码现代化，但在正确性、性能可移植性及可扩展性不容妥协的生产级要求下，仍表现脆弱。我们得出结论，LLM在短期内不太可能取代HPC专家，但有望成为软件开发流程中的强大合作者。其有效部署需要更丰富的数据集、与性能分析及调度器的集成、严格的评估框架，以及确保透明度和可信度的治理结构。因此，AI与HPC的融合应被视为一个长期的共同演化过程，每一次进步都会为重塑科学软件开发带来新的挑战与机遇。
英文摘要
Large Language Models (LLMs) are emerging as promising assistants in High-Performance Computing (HPC), where programming remains complex and expertise-intensive. This survey systematically reviews their application across five categories: code generation, parallelization and optimization, frameworks and architectures, evaluation and benchmarking, and broader challenges. The analysis highlights both opportunities and limitations: while general-purpose LLMs perform reasonably well on serial and OpenMP-like tasks, they fall short in distributed paradigms such as MPI, where correctness and scalability are critical. Domain-specialized models (e.g., HPC-Coder, HPC-GPT, chatHPC) achieve higher accuracy through fine-tuning, curated datasets, and retrieval-augmented generation (RAG), yet their scope remains narrow and their evaluations largely limited to benchmarks or micro-kernels. The broader picture is one of dual potential and fragility: LLMs can lower barriers to entry, accelerate prototyping, and support code modernization, but they remain brittle under production-level requirements where correctness, performance portability, and scaling cannot be compromised. We conclude that LLMs are unlikely to replace HPC experts in the near term but are positioned to become powerful collaborators in the software development pipeline. Their effective deployment will require richer datasets, integration with performance analysis and schedulers, rigorous evaluation frameworks, and governance structures that ensure transparency and trust. The convergence of AI and HPC should therefore be understood as a long-term, co-evolutionary process, where each advance uncovers new challenges and opportunities for reshaping scientific software development.
Journal refFuture Generation Computer Systems, 2026, 108618, ISSN 0167-739X
DOI：10.1016/j.future.2026.108618
