MRI: Acquisition of the Lawrence Supercomputer to Advance Multidisciplinary Research in South Dakota

南达科他州大学（USD）将获得、部署和维护一台集群超级计算机，该计算机将以诺贝尔奖获得者和USD校友E。O.劳伦斯作为一个校园范围内的资源提供给所有USD教师，工作人员，博士后，研究生和本科生以及整个南达科他州的研究人员，劳伦斯集群的主要目标是1）加速科学进步，减少发现时间，2）启用和加速科学成果以前不可能，和3）增加学生参与计算辅助研究。最初，劳伦斯集群将支持北达科他州和南达科他州3个机构的7个部门的12个STEM项目，包括21名教师，26名博士后和307名学生。该系统将支持高能物理、人脑、可再生能源和材料科学等科学领域的多学科研究和研究培训。它将有助于回答相当大的公众利益和社会价值的问题，包括暗物质的性质，以及人类大脑和人类行为之间难以捉摸的联系。此外，在纳米材料的光学性质和有机功能材料的设计方面的进展与广泛的材料科学和工程问题有关。研究结果将为新型材料的高效制备奠定基础，劳伦斯团簇的理论峰值性能将超过60 TFLOPS。该系统架构包括XSEDE兼容软件堆栈、通用计算节点、大内存节点、GPU加速节点、交互式可视化节点、高速InfiniBand互连和高容量并行文件系统。除了传统的命令行界面外，劳伦斯集群还将包括一个基于浏览器的用户门户，用于作业提交和管理。

The University of South Dakota (USD) will acquire, deploy, and maintain a cluster supercomputer to be named after Nobel Laureate and USD alumnus E. O. Lawrence. As a campus-wide resource available to all USD faculty, staff, postdocs, graduate students, and undergraduates as well as researchers across South Dakota, the Lawrence Cluster's key objectives are to 1) Accelerate scientific progress and reduce time to discovery, 2) Enable and accelerate scientific results not previously possible, and 3) Increase student engagement in computationally assisted research. Initially, the Lawrence Cluster will support 12 STEM projects across 7 departments at 3 institutions in North and South Dakota, including 21 faculty, 26 postdocs, and 307 students. The system will support multidisciplinary research and research training in scientific domains such as high energy physics, the human brain, renewable energy, and materials science. It will help answer questions of considerable public interest and societal value, including the nature of dark matter, and the elusive links between the human brain and human behavior. Moreover, advances in both optical properties of nanomaterials and in design of organic functional materials are relevant to a broad range of material science and engineering problems. Results will help lay the foundation for the efficient fabrication of novel materials.The Lawrence Cluster will have a peak theoretical performance of more than 60 TFLOPS. The system architecture includes an XSEDE-compatible software stack, general-purpose compute nodes, large memory nodes, GPU-accelerated nodes, interactive visualization nodes, a high speed InfiniBand interconnect, and a high-capacity parallel filesystem. In additional to a traditional command line interface, the Lawrence Cluster will also include a browser-based user portal for job submission and management.