Description
The purpose of this study was to gain a better understanding of the capabilities of large language models (LLMs), an emerging class of AI algorithms, to benefit the overall HPC community. Key goals of this effort included describing the base of current and planned HPC-related activity that could incorporate LLMs, assessing the level of ongoing LLM activity within end user organizations, characterizing the interest in general-purpose LLM applications, exploring the prospects for LLM integration into traditional HPC algorithms, and highlighting the key challenges with integrating LLM capability into HPC-based workloads.
