CC* Networking Infrastructure: YinzerNet: A Multi-Site Data and AI Driven Research Network

点击翻译按钮获取中文摘要

YinzerNet supports groundbreaking collaborative scientific research and education initiatives between the University of Pittsburgh, Carnegie Mellon University, and the Pittsburgh Supercomputing Center. This network infrastructure eliminates bottlenecks in data movement and sharing and research computing, enhancing the productivity of interdisciplinary collaborations. With its innovative design, YinzerNet integrates modified Data Transfer Nodes at the edge (cloudlets), equipped with Graphics Processing Units (GPUs), Data Processing Units (DPUs), and a basic OpenStack instance, facilitating the deployment of containerized workflows. The network employs Software for Open Networking in the Cloud (SONiC), an open-source network OS overlay, to connect distributed sites, fostering experimental and enterprise network solutions, and adds the ability to expand and connect with other academic institutions in the future.By advancing network capabilities and supporting cloudlet-based distributed research platforms, YinzerNet enables the exploration of new federated Artificial Intelligence (AI) models and network monitoring tools, crucial as federated services emerge as a standard. It serves a wide range of research domains, including satellite communications, security, life sciences, robotics, high-energy physics, astronomy, and neuroscience, demonstrating immediate impacts on these fields.YinzerNet not only represents a significant step forward in networking infrastructure but also acts as a catalyst for future technological advancements and connections to the national cyberinfrastructure ecosystem. Its design and implementation underscore the importance of collaborative platforms in accelerating scientific discovery and innovation, showcasing the potential for significant societal benefits through enhanced research capabilities.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.