RCE an HPC Podcast
Brock Palen and Jeff Squyres speak with Todd Gamblin about Spack. Spack is a package management tool designed to support multiple versions and configurations of software on a wide variety of platforms and environments. It was designed for large supercomputing centers, where many users and application teams share common installations of software on clusters with exotic architectures, using libraries that do not have a standard ABI. Spack is non-destructive: installing a new version does not break existing installations, so many configurations can coexist on the same system.
Todd is a computer scientist in the Center for Applied Scientific Computing at Lawrence Livermore National Laboratory . His research focuses on scalable tools for measuring, analyzing, and visualizing performance the performance of massively parallel simulations. Todd works closely with production simulation teams at LLNL, and he likes to create tools that users can pick up easily.
Frustrated with the complexity of building HPC performance tools, Todd started developing Spack two years ago to allow users to painlessly install software on big machines. Spack has since been adopted by Livermore Computing, other HPC centers, and LLNL application teams. The open source project now includes several core developers at LLNL and a rapidly growing community on GitHub. A 1.0 release is coming soon.
Cyrus is a computer scientist and group leader in the Applications, Simulations, and Quality (ASQ) division of LLNL's Computation directorate. He is the software architect of the VisIt open source visualization tool and leads major aspects of the technical direction of the project. Cyrus also provides custom data analysis solutions for large scale scientific simulations in WCI's WSC and WPD programs.
The Fasterdata Knowledge Base provides proven, operationally sound methods for troubleshooting and solving performance issues. For over 25 years, ESnet has operated an advanced research network with the goal of enabling the highest levels of performance for the Department of Energy (DOE) scientific community. During this time, our engineers have identified a common set of issues that hinder performance and we would like to share our experiences and findings in this knowledge base.
Eli Dart is a network engineer in the ESnet Science Engagement Group, which seeks to use advanced networking to improve scientific productivity and science outcomes for the DOE science facilities, their users, and their collaborators. Eli is a primary advocate for the Science DMZ design pattern, and works with facilities, laboratories, universities, science collaborations, and science programs to deploy data-intensive science infrastructure based on the Science DMZ model. Eli also runs the ESnet network requirements program, which collects, synthesizes, and aggregates the networking needs of the science programs ESnet serves.
Eli has over 15 years of experience in network architecture, design, engineering, performance, and security in scientific and research environments. His primary professional interests are high-performance architectures and effective operational models for networks that support scientific missions, and building collaborations to bring about the effective use of high-performance networks by science projects.
As a member of ESnet's Network Engineering Group, Eli was a primary contributor to the design and deployment of two iterations of the ESnet backbone network - ESnet4 and ESnet5. Prior to ESnet Eli was a lead network engineer at NERSC, DOE's primary supercomputing facility, where he co-led a complete redesign and several years of successful operation of the high-performance network infrastructure there. In addition, Eli spent 14 years as a member of SCinet, the group of volunteers that builds and operates the network for the annual IEEE/ACM Supercomputing conference series, from 1997 through 2010. He served as Network Security Chair for SCinet for the 2000 and 2001 conferences and was a member of the SCinet routing group from 2001 through 2010. Eli holds a Bachelor of Science degree in Computer Science from the Oregon State University College of Engineering.