Dear CASPER Collaboration and Advisory Board Members,
On behalf of the science engagement and engineering teams of the Pacific Research Platform (PRP), we wanted to explore the possibility of a collaboration between the CASPER Collaboration <https://casper.berkeley.edu/index.php/about/> and the PRP in regards to, for example, *the utilization, implementation and development of FPGAs (both hardware and software) for various use cases and applications*, along with other possible areas of collaboration. The PRP is a partnership of more than 50 institutions, led by researchers at UC San Diego and UC Berkeley and includes the National Science Foundation, Department of Energy, and multiple research universities in the US and around the world. The PRP builds on the optical backbone of Pacific Wave, a joint project of CENIC and the Pacific Northwest GigaPOP (PNWGP) to create a seamless research platform that encourages collaboration on a broad range of data-intensive fields and projects. The PRP aims to achieve transparent and rapid data access among collaborating scientists at multiple institutions through an integrated implementation of data-focused networking that extends the university campus Science DMZ model to a regional, <https://pacificresearchplatform.org/NRP/>national, and, eventually, a global scale. PRP researchers are routinely achieving high-performance end-to-end networking from their labs to their collaborators’ labs and data centers, traversing multiple, heterogeneous Science DMZs and wide-area networks connecting multiple campus gateways, enabling researchers across the partnership to transfer data over dedicated optical lightpaths at speeds from 10Gb/s to 100Gb/s. The endpoints in all these PRP sites are devices we call Flash I/O Network Appliances (FIONAs), rack-mounted PCs designed by our PRP team. They are Science DMZ Data Transfer Nodes (DTNs), optimized for 10-100Gbps data transfers and large data storage, and are also capable of holding up to eight GPUs or *FPGA add-in boards* or up to 256TB of disk each. Now entering into its sixth year, the PRP project continues its original proposed mission of working with multi-campus science teams chosen from the fields of *particle physics*, *astronomy*, biomedical sciences, *earth sciences*, *machine learning, and scalable data visualization *to build a researcher-defined and data-focused network whose requirements are driven by direct engagements with sophisticated, cyberinfrastructure-knowledgeable science investigators and network professionals. To learn more about the PRP, please visit our website here <https://pacificresearchplatform.org/>. Please let us know if the PRP science engagement and engineering teams can provide assistance to you or more information with regards to improving network connections and network throughput and data transfer speeds, as well as finding data storage solutions, connecting to the PRP Kubernetes Nautilus Cluster (https://pacificresearchplatform.org/nautilus/), and providing access to CPUs, GPUs, FPGAs, and other computing resources. *Note that **Nautilus also provides access to the latest tensor cores and FPGAs.* Overall, the PRP can assist you in optimizing your workflows. For example, the PRP team can suggest resources to help researchers and scientists optimally use their campus Internet2 (http://www.internet2.edu) network connections, which directly reaches a much larger set of campuses, including supercomputer centers. The PRP can also provide the following consultation and help: - Meet with campus network engineers to analyze the data workflow - Diagnose and suggest fixes for networking bottlenecks that limit network connectivity - Optimize network configurations to move data as fast as possible PRP can also help, for example, by finding an ideal place to host a server with a 10Gbps or better network connection, help design and stand up the server and monitor it, as well as teaching you what you need to know about it and advanced distributed file systems so you have backups, secure access, etc *For more information on PRP FPGAs and the PRP's use of FPGA technology, including XILINX FPGA nodes and XILINX FPGA boards, please see the PRP documentation here <https://pacificresearchplatform.org/admindocs/links/hardware/> and here <https://pacificresearchplatform.org/userdocs/running/special/>. Please note that the PRP Engineering Team is also open to having discussions and exploring possible collaborations around the utilization, implementation, and development of FPGA technology (both hardware and software).* The first step in getting access to the PRP Nautilus GPU/CPU/FPGA cluster is to set up an NRP Matrix account. Matrix is the new tool the Nautilus/NRP community members use to communicate and collaborate. Pointers are available within this link: https://pacificresearchplatform.org/userdocs/start/contact <https://urldefense.com/v3/__https://pacificresearchplatform.org/userdocs/start/contact__;!!Mih3wA!UxEynEspPy70WWSi5se4_r-R_SyVSNLCmULApvQcP7ISBEwPIIMFxusXOibIc9qD$> Within the Matrix account using the Element web client or app, join the available public rooms (hitting the + sign to the right of "Rooms" on the app, for example). "Nautilus General" should be available to join. Individuals can then log into the Nautilus Cluster Portal via the following URL, using "Rochester Institute of Technology" as the CILogon identity provider: https://nautilus.optiputer.net Once logged into the Nautilus Portal initially, individuals are granted initial Guest access. Once joined the Matrix Nautilus General room, individuals can send a message to the room requesting the account be validated. Researchers should mention they're working with you and they're a member of NYSERNet. A researcher who will be the point contact for their team's work on Nautilus can request Namespace Admin privileges as well to manage their environment within Nautilus. Researchers can optionally request JupyterLab access as well. Dimitry Mishin (aka "dimm") is the senior system administrator and research developer for Nautilus - his name will come up in various messages within Matrix, and he may field some of the support requests. For more information about the PRP Nautilus cluster, please also see https://pacificresearchplatform.org/userdocs/start/faq/ . You can join the PRP Listserv here https://mailman.ucsd.edu/mailman/listinfo/prp-l For additional network engineering, storage and computing issues (including the use and development of FPGAs) and for questions on the PRP/NRP in general, co-PIs Tom DeFanti ([email protected]) at UCSD and Camille Crittenden ([email protected]) at UC Berkeley, or senior network engineer John Graham ([email protected]) at UCSD, as well as Applications Developer Dmitry Mishin at San Diego Supercomputer Center / UCSD ([email protected]), will be able to answer questions. Thank you, PRP Science Engagement and Engineering Teams -- You received this message because you are subscribed to the Google Groups "[email protected]" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/CAO%3DeaUQFTsi6FWRjdnf_0C8UjnVPiuFG9LHBQV9Di4Zn5w2Wfg%40mail.gmail.com.
