Speaker: Kate Keahey – Senior Fellow, University of Chicago Computation institute

Abstract: Chameleon is a large-scale, deeply reconfigurable experimental platform built to support Computer Sciences systems research. Kate will will explain the challenges faced in building Chameleon, lessons learned, and operations experiences. She will also describe the packaging of the system that integrates both the developed capabilities and the operational experience and facilitates managing platforms of this kind.

Speaker: Michael Zink – Associate Professor of Electrical and Computer Engineering, University of Massachusetts Amherst

Abstract: A team of researchers at the University of Massachusetts, Boston University, and Northeastern University recently received an award from the National Science Foundation to construct and support a testbed for research and experimentation into new cloud Testbeds such as this are critical for enabling research into new cloud technologies.

Speaker: Ron Hutchins - Vice president for Information Technology, University of Virginia

Abstract: Ron Hutchins is currently VP of IT at University of Virginia in Charlottesville, VA. Ron's focus at UVA has included growing research computing services in both the systems and the support team. Through two NSF awards (CC* and MRI) UVA has focused on high performance connectivity and protected data. The 2019 award of the ACCORD MRI grant is expanding support for computing resources for protected data, both HIPAA and CUI, for UVA and the other public universities in Virginia. Through this award, the ACCORD partners will be able to host protected data and perform high throughput and highly parallel computations on that data. A focus is being put on building a library of containers that can be reused to speed up approval of the computing environment for research. This presentation will include a brief overview of the ACCORD program and answer any questions.

The NSF Mid-Scale is about the design of an instrument to support research. We propose to build an instrument to access or connect researchers and their research instruments, data sources, and associated computational capabilities through federation --, a software and hardware defined “federated collaboratory,” designed to simplify multi-campus collaborations and partnerships that advance the frontiers of research and innovation.

Realizing the vision of the federated collaboratories will require the development of many layers of abstractions ranging from hardware, networking, federation architecture, scientific workflows, and domain-specific models and tools to enable collaborative discovery. In preparation for upcoming workshops focused on gathering information to be included in the ERN Mid-Scale preproposal, this session will be an ERN community forum to discuss what the “federated collaboratory” might look like from both a hardware and software perspective as well as what federation should look like as we strive for a seamless collaborative sharing experience. Another goal of the session is to start identifying an ERN technical team who will lead the architecting/federating/software strategies for the Mid-Scale proposal.

Cryo-EM/ET is one of the research areas where gaining a deeper understanding of the workflows, research computing and data requirements, collaborations, and challenges will enable the ERN to have the broadest impact across multiple research disciplines, pedagogical approaches, senior level college and university administrators, and other organizations within the region and beyond. We estimate that the Cryo-EM/Cryo-ET community in the Northeast comprises nearly 50 centers serving more than 800 laboratories from Pennsylvania to Maine. Applications of Cryo-EM span structural biology and material science. Single Particle Reconstruction information produced by these centers is producing transformative insights. Given the cost and value of the instruments involved, fast turnaround and efficient use of resources is key. While all centers are well equipped to deliver images from prepared samples, processing and storage of these images can present significant and unnecessary obstacles, especially for labs that do not have easy access to computing resources and expertise. The Cryo-EM/Cryo-ET microscopy labs in the Northeast have formed a relatively tight knit community, allowing for free flow of information and experience, and reducing duplication of effort, and accelerating the adoption of new techniques. This session will explore possibilities for extending this collaboration to include the community of Research Computing and Networking organizations that serve these labs and the broader impacts and to form a working group to focus on future workshops and content for the Mid-Scale proposal.

Materials Discovery is one of the research areas where gaining a deeper understanding of the workflows, research computing and data requirements, collaborations, and challenges will enable the ERN to have the broadest impact across multiple research disciplines, pedagogical approaches, senior level college and university administrators, and other organizations within the region and beyond. Researchers in materials discovery are realizing that their traditional data-intensive HPC workflows are reaching the limits of spatial and temporal scales required to make deeper insights and predictions. For this reason, they are looking to new paradigms that include convergence of HPC and Machine Learning (ML) methodologies, algorithm development, and novel ways to access the data distributed across multiple institutions used in training systems as promising approaches to overcome the major computational performance limitations. Materials Discovery offers an attractive testbed for advanced cyberinfrastructure of the sort the ERN can offer through future funding opportunities such as the Mid-Scale RI-1 program and DMREF. As with Cryo-EM/Cryo-ET, this session will explore possibilities for extending collaborations to include other institutions as well as the community of Research Computing and Networking organizations and to form a working group to focus on future workshops and content for the Mid-Scale proposal.

The vision of the ERN is to simplify multi-campus collaborations and partnerships that advance the frontiers of research and innovation. In order to do this successfully the ERN needs to consider current university policies as well as engage with university administrations (VPRs, CIOs, General Counsel, and IRB directors) in developing a policy strategy to help us bring the vision to reality. This session will give the ERN community a chance to discuss university policies and what considerations need to be taken into account as we create new policies and procedures. What needs to be in place that allows ease of sharing knowledge, data, infrastructure, and people? Sustainability, another important discussion point, as well as compliance requirements and security concerns should also be addressed during this session. A goal of this session is to create a working group focused on CI sharing policies as well as creating content to be included in the Mid-Scale proposal.

Many smaller, mid-sized and under-resourced campuses, including MSIs, HSIs, HBCUs, and EPSCoR institutions, have compelling science research and education activities along with an awareness of the benefits associated with better access to cyberinfrastructure resources. These schools can benefit greatly from resources and expertise to augment their in-house efforts. This could include identifying, understanding, and quantifying the science drivers; understanding the cyberinfrastructure needed to support the applications; and providing both the technical and application support associated with matching the applications to the infrastructure, particularly when the required resources are outside of their campus environment. This session is focused on learning directly from this community on how best to support under-resourced academic institutions in the region, which happens to be a majority of the academic institutions within the Northeast. Because of the trusted relationship between smaller academic institutions and the regional network providers, the role of regionals as facilitator and user support for these smaller institutions within the ERN will also be explored during this session. A goal of this session is to form a working group focused on broadening the reach as well as creating content to be included in the Mid-Scale proposal.

We believe that the ERN will both benefit from and serve as a testbed for Computer science research. As we strive to develop solutions focused on supporting the vision and mission of the ERN, many of the challenges the ERN will face will lead to questions that ultimately become interesting CS research projects on topics such as ontologies and knowledge representation, workflow analysis, federated AI/ML, domain-specific programming languages, to architecture, networks, systems, and security. The ERN will also become a testbed for CS experiments (Measurement Monitoring, Self Optimizing Systems) and will leverage existing NSF funded projects (FABRIC, Open Cloud, CC*) leading the way to become an instrument to connect other research instruments or platforms. In this session the focus will be on how computer science research can play an important role in defining the ERN and to form a working group to focus on the different CS related projects to be included in the Mid-Scale proposal.

Breakout session leads will give a high level overview of what was discussed in their respective sessions and the working groups formed.

The vision of the federated collaboratories will require the development of many layers of abstractions ranging from hardware, networking, federation architecture, scientific workflows, and domain-specific models and tools to enable collaborative discovery. The ERN Architecture and Federation working group are focused on gathering information and discussing what the “federated collaboratory” might look like from both a hardware and software perspective as well as what federation should look like as we strive for a seamless collaborative sharing experience.

As a member of the ERN community, or attendee of the ERN All Hands Meeting in July, we invite you to participate in the Eastern Regional Network’s Architecture and Federation workshop December 2nd - 4th. The workshop is currently scheduled to cover the following topics:

-Authentication, authorization and access

-Data storage and transfer

-Network connectivity

-Federated infrastructures science driver use cases

Discussions following each session’s presentations will help in the understanding of how best to model the federated instrument promoting collaboration and scientific
advancement.

Please click on the event title to learn how to register and to view the workshop schedule.