CC* Regional Computing: Great Plains Extended Network of GPUs for Interactive Experimenters (GP-ENGINE)

GP-ENGINE项目推动了在大平原网络地区采用先进的计算和数据资源。该项目将增加地方和国家计算资源服务的研究人员和学生的数量，加强校园研究计算专业人员的能力和能力，并扩大区域研究能力。研究人员将能够使用本地提供的高级计算资源将新生的想法和代码转换为高级计算代码。这些代码可以稍后在国家高吞吐量计算资源上执行。这些成功将加强机构对可持续的区域和国家研究计算系统的支持。该项目利用强大的现有合作来提供和管理密苏里州、北达科他州、南达科他州、堪萨斯州、俄克拉何马州和阿肯色州的图形处理单元资源。它训练和支持研究人员将他们的工作台代码改编成高吞吐量的计算代码，这些代码可以在国家平台上执行，如开放科学网格和国家研究平台。该项目解决了来自整个联盟的各种科学驱动程序的计算需求，包括3D蛋白质分子生成、野火烧伤区域测绘的卫星图像深度学习、暗物质和中微子探测、陆地表面物候的实时监测、网络安全攻击图生成以及数字孪生的智能制造等问题。该项目通过采用JupyterLab计算机笔记本来改善STEM研究和教育。部署的网络基础设施支持该地区的研究和教育，该地区人口稀少，地理面积大，非常适合在资源不足的EPSCoR州开发先进的计算、数据和网络。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The GP-ENGINE project advances the adoption of advanced computing and data resources in the Great Plains Network region. This project will increase the number of researchers and students served by both local and national computing resources, strengthen the capacity and capabilities of campus research computing professionals, and expand the regional capacity for research. Researchers will be able to transition nascent ideas and codes into advanced computing code using locally provisioned advanced computing resources. These codes can be later executed on national high-throughput computing resources. These successes will enhance institutional buy-in for sustainable regional and national research computing systems.The project leverages strong existing collaborations to provision and manage graphics processing unit resources in Missouri, North Dakota, South Dakota, Kansas, Oklahoma, and Arkansas. It trains and supports researchers to adapt their workbench codes into high-throughput computing codes that can be executed on national platforms such as the Open Science Grid and the National Research Platform. The project addresses computing needs of a diverse set of science drivers from across the consortium, including problems in 3D protein molecule generation, satellite image deep learning for wildfire burn area mapping, dark matter and neutrino detection, real-time monitoring of land surface phenology, cybersecurity attack graph generation, and intelligent manufacturing with digital twins. The project improves STEM research and education through adoption of JupyterLab computing notebooks. The deployed cyberinfrastructure supports research and education in a region whose sparse population and geographic size are ideal for developing advanced computing, data, and networking in under-resourced EPSCoR states.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.