2017 PBS Works User Group Speakers
Senior Vice President, Global Enterprise Solutions, Altair
With over 20 years of simulation experience ranging from application engineering, large account management, regional director, SVP Americas, to his current role as the Senior Vice President-Global Enterprise Solutions, Joe Sorovetz is responsible for the global business unit which includes all field operations, including technical support, marketing, and administration throughout the globe.
Having been at Altair since 1999, Joe has been involved with all of the various verticals and business units within Altair. Joe holds a Bachelor's Degree in Mechanical Engineering from the University of Michigan-Dearborn, graduating in 1991.
Chief Technical Officer, HPC/Cloud Solutions, Altair
Srikanth (Sam) Mahalingam is the Chief Technical Officer for Altair’s cloud computing and high-performance computing strategy, with the primary responsibility of applying the newer cloud and mobile technologies to the digital simulation, high performance computing world and identifying and mentoring talented individuals within the organization.
With more than 20 years of experience in software development, software architecture, technical management and project management, Mahalingam’s role focuses on shaping the current products and identifying newer products and solutions to ease the cloud adoption and mobile strategies for Altair customers, both in the Simulation Lifecycle Management and High Performance Computing Lifecycle Management arenas.
Mahalingam joined Altair in 1994 and worked on various products and projects in both the custom services and product teams. He then joined a technology start-up as a joint promoter for launching a disruptive virtual try-on product into the market that was later acquired by a large software services firm in India. He rejoined Altair in 2005 to spearhead newer product initiatives that include the cloud platform, and eventually became the thought leader within Altair for the Enterprise Software Strategy.
Mahalingam holds a Bachelor of Engineering in Computer Science and Engineering and Executive Master of Business Administration from the Indian School of Business.
Dr. Bill Nitzberg, Moderator
PBS Works CTO
Dr. Bill Nitzberg is the CTO of PBS Works at Altair Engineering, Inc. With over 30 years in the computer industry, spanning commercial software development to high-performance computing research, Dr. Nitzberg is an internationally recognized expert in parallel and distributed computing. Dr. Nitzberg has served on the board of the Open Grid Forum, co-architected NASA’s Information Power Grid, edited the MPI-2 I/O standard, and has published numerous papers on distributed shared memory, parallel I/O, PC clustering, job scheduling, and cloud computing. In his spare time, Bill tries to reduce his pack weight for his long- distance hiking trips.
EVP & Chief Information Officer, Caesars Palace Entertainment
Manager of User Services at the National Computational Infrastructure (NCI)
Roger Edberg is the Manager of User Services at the National Computational Infrastructure (NCI) facility in Canberra, Australia. NCI is the leading provider of high performance computing to scientists in Australia’s academic and government research sectors. At present, NCI provides HPC services to approximately 4600 researchers. Roger earned a PhD in Physical and Theoretical Chemistry from the Australian National University, and has worked in HPC application development, scientific data visualization, development of animation, rendering, and production pipelines for motion picture production, and petabyte-scale calibration of Earth observation data prior to his NCI role.
PBS Professional at NCI
PBS Professional is used on NCI’s petascale computing system, Raijin. Raijin has 4600+ active users across the full spectrum of scientific disciplines from academic, government, and commercial stakeholders in Australia. This presentation will provide an overview of how PBS Professional is used to support scientific processing workloads on Raijin. The presentation will also include a discussion of current NCI resource accounting using PBS Professional, as well as plans to improve and automate reporting and billing processes for NCI stakeholders.
On-site Contractor at Ames Research Center, NASA Ames
Greg Matthews is an on-site contractor at Ames Research Center in the NASA Advanced Supercomputing (NAS) Division, working for CSRA. He has a B.S. in computer science from Harvey Mudd College, where he developed a love for parallel programming and HPC in general. NAS has indulged this love for the past 17+ years with opportunities such as parallel and automated debugging research, supercomputer monitoring, accounting and usage analysis, and PBS Pro administration and development.
Prototype Features for PBSPro at NAS
NASA Advanced Supercomputing (NAS) at NASA Ames uses PBS Pro to schedule and manage the workload on several supercomputing resources, including Pleiades (11,000+ node IB cluster) and Electra (1,152 node IB cluster). NAS works closely with Altair to design and prototype features that help NAS but also apply as generally as possible to other sites running PBS Pro. We'll present recent features that Altair and NAS have either put into production use at NAS, tested in prototype form, or begun design work on. These features affect a wide range of PBS Pro operations, from scheduling performance to admin workflow to user workflow.
HPC Principal Technologist, GE
Thomas Leung is the Principal Technologist of the GE Engineering Compute Shared Service (ECSS) and member of the Enterprise Architect team at GE Global Research Center (GEGR). Thomas has over 15 years of experience in the IT industry, specializing in Supercomputing, HPC, Global Information Technology, and Big Data. In his role at ECSS, he is responsible for architecting HPC and Big Data systems, forming HPC technology strategy, providing system architecture/solutions, and driving HPC technical decisions for GE locations around the world. Thomas leads the @scale team on technical direction around OpenStack Cloud Foundry, Docker, and Machine Learning service offerings. Mr. Leung is an active member of several user groups and an HPC product advisory board. Prior to joining GE, Thomas was the HPC project manager at Knolls Atomic Power Laboratory, where he managed HPC projects and brought in multiple Top500 class supercomputers. Mr. Leung holds an M.B.A. from the State University of New York at Albany, and an M.S. in Information Technology from Rensselaer Polytechnic Institute, New York.
Containers and Orchestration Tool Interaction with HPC Resource Manager
HPC systems stay with leading-edge technology and most challenging simulation workload all the time. There are multiple important technology changes in the HPC market that we explored. For example, container technology is a hot topic in these two years. This will enhance the productivity of HPC systems by offering portability, scalability, and speed of software deployment to HPC users. Not only are there technology changes, but machine learning, analytical workloads, and big data workloads are new to HPC systems as well. Since there are new workloads and new technology, different problems that we need to address such as not being able to directly deploy workloads without orchestration tools or platform-as-a-service tools.
In this presentation, we will explore the challenges introduced by new platforms and new workloads. Understanding the difficulties while integrating them into HPC systems. Discuss about the direction of convergence of computational platforms like HPC systems, Big Data clusters, Openstack and Docker clusters. We will also do a short briefing on Docker integration work done with PBS Pro.
Director of Business Development, HPC Platform Software, Datacenter Group, Intel Corporation
Tom Krueger currently serves as Director of Business Development for the HPC Platform Software team in the Datacenter Group within Intel Corporation. He most recently led the launch of Intel® HPC Orchestrator at SC ‘16. Krueger has over 25 years in technical sales and marketing experience including HPC Software, Embedded Products, Enterprise Servers, Communications Infrastructure and Digital Watermarking technology. He has worked in various roles with Intel for nine years working with End Users, OEMs, SIs and ISVs. Prior to Intel Krueger held executive positions with Experis, Digimarc and Corvalent Corporation in Sales, Marketing and Business Development capacities. Krueger holds a BS degree in Business Administration from Cal Poly Pomona and an MA in Organizational Management from The University of Phoenix. In his spare time, he enjoys assisting start-up companies, hiking, biking, wine tasting and real estate.
Simplifying your HPC System Software Stack with Intel® HPC Orchestrator & PBS Professional
The Intel® HPC Orchestrator software stack is part of the Intel Scalable System Framework and helps enable other Intel products such as the Intel® Xeon Phi™ and Omni-Path Architecture to work better together. This software stack includes integrated components such as PBS Professional that add value to HPC Orchestrator that differentiate it from other HPC software stack solutions. This session will provide an overview of how HPC Orchestrator with PBS allows for robust scheduling, optimal system utilization, and simplified integration, installation, and ongoing maintenance of HPC systems. Intel’s experience in using HPC Orchestrator with PBS Professional on several large Intel lab clusters will also be highlighted.
Senior Director of Oracle’s Bare Metal Cloud Services (BMCS), Oracle
Bruce Burns is the senior director of Oracle’s Bare Metal Cloud Services (BMCS) product management team. His work includes driving technical collaboration with customers and partner teams to help enable them to use the Oracle cloud platform and ensure that the new cloud services and features being built fully meet the needs of the partners and customers that use them. Before joining Oracle, Burns was a senior manager at Amazon where he created the AWS Test Drive and Quick Start programs and ran the AWS global Oracle and SAP partner relationships. Previous to AWS, Burns was a principal program manager at Microsoft with over 16 years of experience building out Windows client, server, and cloud partner ecosystems.
Using Oracle Bare Metal Cloud Services with PBS Professional
Time is money, and in this session you will learn how to save both using PBS works with Oracle Bare Metal Cloud Services (BMCS) public cloud infrastructure. Oracle BMCS is a new generation of scalable, inexpensive and performant compute, network and storage infrastructure that combines internet cloud scale architecture with enterprise scale-up bare metal capabilities, providing the ideal platform for demanding High Performance Computing workloads. In this session you learn how to use PBS Professional with Oracle BMCS for common customer use case scenarios to dynamically scale out, on demand and leverage the price / performance advantages that Oracle BMCS uniquely provides with PBS Works.
Senior HPC Analyst, Boeing
Jim Glidewell has been supporting High Performance Computing systems at Boeing since the early eighties. As a member of the support team for Boeing’s Cray systems, he worked on five generations of Cray systems, starting with the Cray-1. Jim has served as his group’s primary focal for job scheduling and resource accounting for over twenty years. For the past several years, he has served as technical lead of the Enterprise HPC group. He also serves as the technical focal for PBS Pro within Boeing, as well as being responsible for support of PBS Pro on Boeing’s Enterprise HPC Service.
Optimizing HPC Service Delivery
There are multiple elements to providing an effective and efficient HPC service. This presentation will share some of our strategies for extracting maximal value from our HPC hardware and providing a service that meets the needs of our engineering customers.
Senior Manager – Product Specialist, Altair
Dario Dorella started his career as a system administrator back in 1998. While helping manufacturing companies evolving their digital infrastructures to incorporate the emerging HPC technologies in their workflow, Mr. Dorella gathered competencies ranging from network to security and application management. He worked with all the major operating systems (IRIX, HP-UX, Solaris, Linux, Windows) and platforms and taught training classes for SGI IRIX, LSF and PBS Professional. While collaborating with Altair Engineering, he had the opportunity to work with major computing centers and universities around the world to train their personnel and design and deploy HPC solutions based on Altair tools and technologies. He also covered PBS Works Second Level support for EMEA region and continued to work on large accounts addressing pre-sales, delivery and post-sales tasks.
Better Experience for HPC Applications
Thinking outside the browser — PBS Access desktop provides a richer, more streamlined job submission and monitoring for Engineers and Scientists.
Deputy Program Manager for Technology for Team SAIC/HITS-U
William A. Ward, Jr. is Team SAIC/HITS-U Deputy Program Manager for Technology for the High Performance Computing Modernization Program. His technical interests include computer performance evaluation and scientific computing. Ward has a PhD in computer science from Purdue University.
Experimental Determination of Expansion Factor Versus Utilization Using Scheduler Simulation
Historically, HPC systems have been used as heavily as possible in order to justify their expense, thus resulting in high expansion factors. This study uses scheduler simulation to obtain a more precise understanding of the relationship between expansion factor and utilization to allow centers to estimate how much additional resource would need to be supplied to obtain a specified level of user satisfaction.
UNIX system administrator
Using PBS since 2003 – first with Open PBS and since 2005 PBS Pro.
2005 - Present. IT Professional with Chevron Energy Technology Computer, Advanced Computing Unit, responsible mainly for cluster computing and other HPC activities. Heavily involved in all PBS in Chevron.
1999 – 2005 - Unix / Linux Systems Administrator at Unocal, Houston
1991 – 1999 – Seismic Systems Engineer with Schlumberger, Worldwide.
1991 Graduated from University Wales, UK, BSc (Hons) Electronic Engineering.
PBS Pro – Periodic Hook to Build / Destroy Cloud Instances on Demand Based on Queued Jobs
Discuss the proof of concept built and tested in Chevron that utilizes PBS Pro and a periodic server hook to directly query the queues and nodes already built to kick off building and configuring of node instances on demand in AWS cloud to allow queued jobs to run. Nodes are destroyed on completion of the job – again by queue and node monitoring.
Associate Vice President for Economic Development, Executive Director, UNLV Research Foundation
Zach Miles serves as the Associate Vice President for Economic Development. Miles also serves as the Director of the Nevada Small Business Development Center (SBDC) at UNLV and Executive Director for the UNLV Research Foundation. In these capacities, Miles manages the assessment, protection, and commercialization of intellectual property owned or controlled by UNLV and assists start-up companies in building a strong business foundation and seeking capital. Prior to joining UNLV, Miles served as the deputy executive director of Technology Venture and Commercialization at the University of Utah. Miles has also co-founded, advised, and served as general counsel for a number of Utah based start-up companies. Miles holds a Bachelor of Science degree from Weber State University, where he majored in microbiology; he also holds a Juris Doctorate from the University of New Hampshire School of Law.
Supercomputers, Economic Development, and UNLV, oh my!
This presentation will take a look at the unified economic development approach of UNLV; providing a single point of contact for engagement with the community for commercialization of life changing products and services. UNLV's supercomputer, metal rebel, will be discussed as one of the tools UNLV has to engage with the community in economic development and diversification.
Jignesh (Jig) Bhadaliya,
CTO, HPC/EDA, Dell EMC
Jig is the CTO for HPC/EDA at Unstructured Storage Division. Jig has over twenty years of experience in HPC/EDA, having worked at Dell EMC, Broadcom, Rockwell Semiconductors and University of California. Jig has been working in high performance computing and storage for over a decade, and is passionate about quantifying, characterizing, synthesizing, and replaying workloads to understand their effect on the overall HPC infrastructure. His most recent contribution defines the SFS2014 SP2 EDA benchmark suite, to be released later this year. Jig is based out of Orange County, California, where he has spent the majority of his life. When not working, Jig enjoys spending time with his family, bicycling, or hacking embedded devices.
Storage Aware Job Scheduling
In any HPC environment, storage is an integral and often unaccounted performance component of the infrastructure alongside compute. The health and performance of the overall infrastructure can be highly dependent on storage. Storage related performance issues can lead to adverse effects on overall job throughput, often resulting in business impact. In a reactive manner, the issue is often addressed by throttling the job flow. A more elegant and proactive approach would be to establish a feedback loop between storage and compute with the assistance of a job scheduler. This talk will explore the need for such a feedback loop, its benefits, and potential implementation scenarios for embarrassingly parallel HPC workloads.
Propulsions Systems Division, Orbital ATK
Ramesh Krishnan is a senior staff engineer and manages the corporate HPC environment at Orbital ATK, an Aerospace and Defense manufacturing company (orbitalatk.com). His responsibilities include system, software & storage evaluations, capacity planning, RFP preparation and technology evaluations, software migration and performance testing, improving system utilization & efficiency, and on-going support & governance. In addition, he oversees design & analysis (CAE) tools used across the Enterprise, driving commonality and process improvements. His activities in this role include needs assessments, technology evaluation & deployment, usage monitoring, and ISV liaison. He holds a BS in Civil Engineering from the Indian Institute of Technology, and an MS in Structural Engineering from the Univ. of Cincinnati.
Consolidating HPC Systems & Services: Challenges and Benefits
Orbital ATK is an Aerospace & Defense Manufacturing & Technologies company, headquartered in Dulles, Virginia, with approximately 12,500 employees in 18 states across the U.S. Orbital ATK designs, builds and delivers space, defense and aviation-related systems to customers around the world both as a prime contractor and as a merchant supplier. Our main products include launch vehicles and related propulsion systems; satellites and associated components and services; composite aerospace structures; tactical missiles, subsystems and defense electronics; and precision weapons, armament systems and ammunition.
Orbital ATK has chosen to consolidate and centralize its HPC resources and services, currently in Promontory, Utah, for all its divisions across the US. This decision brings with it both technical and cultural challenges, as well as significant benefits to the business. This presentation will address both aspects and lessons learned along the way.
Jérémie Bourdoncle MSc, MBA,
Director, PBS Cloud, Altair
Jérémie Bourdoncle joined Altair in 2015 as director for PBS Cloud; he oversees the product management and business development. Prior to joining Altair, Jérémie was the co-founder and CEO of Hedera Technology for 8 years. The start-up has developed an innovative Cloud Management Platform that automates and orchestrates application deployment on AWS and OpenStack. Jérémie holds a Master of Engineering degree in Computer Science from Telecom ParisTech and Executive Master of Business Administration from EMLyon.
PBSCloud.io – Get your HPC Cloud Under Control
Cloud Computing opens a vast ocean of opportunity for HPC users. But all these opportunities come with risks: being locked in, adding more IT complexity, budget explosion… PBScloud.io is a new HPC cloud management platform that puts cloud infrastructures under your control.
With PBSCloud.io, you can:
- Avoid cloud provider lock-in by choosing between major global and local cloud providers, and change from one to another simply
- Create your own HPC appliance by designing your HPC infrastructure with a simple and intuitive interface
- Integrate your application with PBS Works, our world-leading workload scheduling management suite
- Deploy your HPC stack everywhere in minutes with a few clicks
- Control your cloud consumption with user quotas and alerting
With PBSCloud.io you can model, create, manage and supervise your HPC appliance.
President, insideHPC Media
Bio: Recently named as one of the Top 20 Big Data Influencers by Forbes Magazine, Rich Brueckner is an avid writer, publisher, and technology pundit focused on high performance computing. He acquired inside-HPC.com in 2010 and has since expanded his online publications to include inside-BigData, inside-Startups, and The Exascale Report. With over 25 years of HPC experience at Cray Research, SGI, and Sun Microsystems, Rich is known to many in the industry as “the guy in the Red Hat.”
insideHPC: Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. Written and edited by supercomputing professionals with the help of readers and occasional contributors, insideHPC sifts through all the news so you don’t have to!
Manager, WW Storage Systems Engineering
Bio: Dale Brantly leads the global Panasas systems engineering group responsible for architecting and delivering ActiveStor storage solutions to customers worldwide.
Dale has more than 30 years of experience in defining engineering requirements and achieving architectural and quality excellence for complex storage solutions deployed in a large high performance computing and enterprise environments.
Prior to joining Panasas, Dale managed global solutions development at Silicon Graphics (SGI) and ran SGI’s Information Lifecycle Management (ILM) product line. Other appointments include field and customer site technical positions at Cray Research, Control Data Corporation, and ETA Systems.