KAIST网络中间件研究中心之网络与计算实验室正在研究协同高性能计算互联、基于深度学习的平台、集成中间件管理计算资源等。高性能计算（HPC）和分布式计算（DC）系统已经逐渐普及。现在，任何可计算的计算机（包括从集群到互联网工作的计算机和移动设备）都需要高效计算以支持各种应用程序。因此，该研究所的工作主要集中在高性能和分布式计算系统以及计算中间件的设计和工程背后的关键技术。
研究所正在进行很多新的研究项目，基于数据映射的大数据智能搜索与利用开发方面，基于各种数据中心的互操作性，自动配置数据映射，实时反映数据集的链接和变化，开发智能大数据搜索和利用核心技术，支持用户分析的精确数据集搜索和利用。
关于5G的低延迟设备问题，为了改善混合现实中的用户体验，低延迟集成了信息处理与5G边缘云计算和多租户交互调度技术。
分析加速基于XAI的卫星图像分析加速，通过机载系统设计并加速卫星的自适应分析模型，将分析后的图像信息连续提供给地面数据中心。
用于医疗数据分析的自改进集成人工智能系统，自适应代理提取自己的高级特征，以帮助根据当前遇到的状态数据做出决策。
The Network and Computing Lab of KAST Network Middleware Research Center is researching collaborative high-performance computing interconnection, deep learning-based platform, integrated middleware management computing resources, etc. High-performance computing (HPC) and distributed computing (DC) systems have become increasingly popular. Now, any computable computer, including computers and mobile devices working from clusters to the Internet, needs efficient computing to support a wide range of applications. Therefore, the institute's work focuses on the key technologies behind the design and engineering of high-performance and distributed computing systems and computing middleware.
The Institute is carrying out many new research projects, based on the development of big data intelligent search and utilization based on data mapping, based on the interoperability of various data centers, automatic configuration of data mapping, real-time reflection of data set links and changes, development of intelligent big data search and utilization core technology, support user analysis of accurate data set search and utilization.
Regarding the low-latency device issue of 5G, in order to improve the user experience in mixed reality, low-latency integrates information processing with 5G edge cloud computing and multi-tenant interactive scheduling technology.
Analysis acceleration Based on XAI's satellite image analysis acceleration, the analyzed image information is continuously provided to the ground data center through the airborne system design and acceleration of the satellite's adaptive analysis model.
A self-improving integrated AI system for medical data analysis, adaptive agents extract their own high-level features to help make decisions based on currently encountered state data.