SEHET22

Fifth Workshop on Strategies for Enhancing HPC Education and Training (SEHET22)

Monday July 11, 2022
Time: 8:30 AM - 11:30 AM (EST)

08:30 AM-8:45 AM

Welcome SIGHPC Education Chapter

Nitin Sukhija, Slippery Rock University of Pennsylvania

8:45 AM -10:45 AM

Paper Session

8:45 AM -9:00 AM

“The Multi-Tier Assistance, Training, and Computational Help (MATCH) project, a Track 2 NSF ACCESS Initiative”. Shelley Knuth, Julie Ma, Alan Chalker, Ewa Deelman, Layla Freeborn, VikrAM Gazula, John Goodhue, JAMes Griffioen, David Hudak, Andrew Pasquale and Mats Rynge

NSF-supported cyberinfrastructure (CI) has been highly successful in advancing science and engineering over the last few decades. During that time, there have been significant changes in the size and composition of the participating community, the architecture and capacity of compute, storage, and networking platforms, and the methods by which researchers and CI professionals communicate. These changes require rethinking the role of research support services and how they are delivered. To address these changes and support an expanding community, MATCH is implementing a model for research support services in ACCESS that comprises three major themes: 1) leverage modern information delivery systems and simplify user interfaces to provide cost-effective, scalable, support to a broader community of researchers, 2) engage experts from the community to develop training materials and instructions that can drAMatically reduce the learning curve, and 3) employ a matchmaking service that will maintain a database of specialist mentors and student mentees that can be matched with projects to provide the domain-specific expertise needed to leverage ACCESS resources. A new ACCESS MATCH Portal (AMP) will serve as the single front door for researchers to obtain guided support and assistance. AMP will leverage emerging curated and tagged taxonomies to identify and match inquiries with knowledge base content and expertise. Expert-monitored question and answer platforms will be created to ensure researcher questions are accurately answered and addressed in a timely fashion, and easy-to-use interfaces such as Open OnDemand and Pegasus will be enhanced to simplify CI use and provide context-aware directed help. The result will be a multi-level support infrastructure capable of scaling to serve a growing research community with increasingly specialized support needs, resulting in research discoveries previously hindered by researchers’ inability to effectively utilize NSF CI resources. This talk will cover the components of the MATCH project and discuss how MATCH will engage and work with the ACCESS community.

9:00 AM -9:15 AM

“Enhancing HPC Education and Workflows with Novel Computing Architectures”. Jeffrey Young, Semir Sarajlic, Aaron Jezghani, WilliAM Powell, Jeffrey Valdez, Michael D. Weiner, SAM Jijina and Xueyang Liu

Recent HPC education efforts have focused on maximizing the usage of traditional- and cloud-based computing infrastructures that primarily support CPU or GPU hardware. However, recent innovations in CPU architectures from Arm and RISC-V and the acquisition of Field-ProgrAMmable Gate Array (FPGA) companies by vendors like Intel and AMD mean that traditional HPC clusters are rapidly becoming more heterogeneous. This work investigates one such exAMple deployed at Georgia Tech - a joint workflow for processor design and reconfigurable computing courses supported by both the HPC-focused Partnership for Advanced Computing Environments (PACE) and GT’s novel architecture center, CRNCH. This collaborative workflow of HPC nodes and 40 remotely accessible Pynq devices supported over 100 students in Spring 2022, and its deployment provides key lessons on sticking points and opportunities for combined HPC and novel architecture workflows.

9:15 AM -9:30 AM

“Developing Standards for HPC Education and Training Repositories”. Susan Mehringer, Kate Cahill, Jp Navarro, David Joiner, Aaron Weeden and Scott Lathrop
During the Education and Training Workshop at PEARC22, we plan to discuss the needs, requirements and strategies for building and implementing repositories that can be shared and exchanged AMong a diverse mix of institutions on a national and international scale. This will include working to create common metadata for publishing and exchanging training and educational material to facilitate common indexing and searching methods. To continue and build on the workshop presentation and discussion, we plan to propose the formation of a Repository Working Group within the ACM SIGHPC Education Chapter. The use of computing technologies is expanding exponentially in every sector of our lives, creating a need for access to high-quality education and training materials to conduct research computing. Instructional materials for both teaching and learning are needed on a broad range of topics related to both developing and applying research computing technologies in all disciplines. The critical need for quality materials applies to formal classroom learning as well as to informal and self-paced learning. Across the globe, academic, industrial, and government organizations develop local web-based repositories (i.e. portals or collections) where their stakeholders can access training and educational materials. However, organizations do not have common standards or protocols for publishing and discovering high-quality learning materials via their repositories. While most people still use search engines like Google to discover materials, we believe there is a vital need for the curation of education and training repositories. Curated repositories can provide verification, validation, and accreditation information, reviews by peer groups, and roadmaps for learning. One of the greatest benefits of common curation methods is providing easier access to materials that are high-quality, up-to-date, maintained, and relevant to people’s background and needs. We believe that by promoting common strategies for structuring and indexing repositories, we can reduce the time and effort by people publishing and/or discovering quality HPC education and training materials.

9:30 AM -9:45 AM

“HPC Workforce DeveloAMent of Undergraduates Outside the R1”. Scott Feister and Elizabeth Blackwood
Many Research-1 (R1) universities create investments in High Performance Computing (HPC) centers to facilitate grant-funded computing projects, leading to student training and outreach on cAMpus. However, creating an HPC workforce pipeline for undergraduates at non-research-intensive universities requires creative, zerocost education and exposure to HPC. We describe our approach to providing HPC education and opportunities for students at California State University Channel Islands, a four-year university / Hispanic-Serving Institution (HSI) with a primarily first-generationto-college LatinX student population. We describe how we educate our university population in HPC without a dedicated HPC training budget. We achieve this by (1) integrating HPC topics and projects into non-HPC coursework, (2) organizing a cAMpus-wide data analysis and visualization student competition with corporate sponsorship, (3) fielding undergraduate teAMs in an external, equity-focused supercomputing competition, and (4) welcoming undergraduates into faculty HPC research. The net effect of this multifaceted approach is that our graduates are equipped with core competencies in HPC and are excited about entering HPC careers.

9:45 AM -10:00 AM

“Bridging Data Science ProgrAMming with Advanced Formal Coursework”. Wesley A. Brashear, Zhenhua He, Richard Lawrence, Dhruva K. Chakravorty, Tatevik Sekhposyan, Margaret L. Carpenter and Honggao Liu
In order to fulfill the needs of an evolving job market, formal academic progrAMs are continuously expanding computational training in traditional discipline-specific courses. We developed an informal, twelve contact-hour course tailored for economics students entering a computationally rigorous graduate-level course to help mitigate disparities in computing knowledge between students and prepare them for more advanced instruction within the formal setting. The course was developed to teach the R progrAMming language to students without assuming any prior knowledge or experience in progrAMming or the R environment. In order to allow for ease of implementation across various training approaches, the course is modularized with each section containing distinct topics and learning objectives. These modules can be easily developed as independent lessons so that discipline-specific needs can be addressed through inclusion or exclusion of certain topics. This implementation used the R package ‘learnr’ to develop the course which rendered a highly extensible and portable interactive Shiny document that can be deployed on any system on which RStudio is installed. The course is offered as a series of interactive sessions during which students are led through the Shiny notebook by an instructor. Owing to its structure, it can be offered as an asynchronous web-based set of tutorials as well.

10:00 AM -10:30 AM

Break

10:30 AM -10:45 AM

" Exascale Computing Project’s Broadening Participation Initiative". Ashely Barker and Mary Ann Leung
This talk will give an overview of the Exascale Computing Project’s Broadening Participation Initiative, which has the mission of establishing a sustainable plan to recruit and retain a diverse workforce in the DOE high-performance computing (HPC) community by fostering a supportive and inclusive culture within the computing sciences at DOE national laboratories. We describe key activities within three complementary thrusts: establishing a Workforce DeveloAMent and Retention Action Group, creating accessible ‘Intro to HPC’ training materials, and launching the Sustainable Research Pathways for High-Performance Computing (SRP-HPC) workforce develoAMent progrAM. We are leveraging ECP’s unique multilab partnership to work toward sustainable collaboration across the DOE community, with the long-term goal of changing the culture and demographic profile of DOE computing sciences.

10:45 AM -11:30 AM

Discussion Session

10:45 AM -11:25 AM

Panel: “Cloud / HPC clusters for Accelerating Training and Education”
Charlie Dey (Texas Advanced Computing Center), Mary Thomas (San Deigo Supercomputer Center) and Henry Neeman (University of Oklahoma)

11:25 AM -11:30 AM

Next Steps and Action Plans