News

SCA20 Speakers

Keynote Speakers

Director of IBM Research

Dr Dario Gil is the Director of IBM Research, one of the world’s largest and most influential
corporate research labs. He is the 12th Director in its 74-year history.

IBM Research is a global organization with over 3,000 researchers at 19 locations on six
continents advancing the future of computing. Dr Gil leads innovation efforts at IBM, directing
research strategies in Quantum, AI, Hybrid Cloud, Security, Industry Solutions, and
Semiconductors and Systems.

Prior to his current appointment, Dr Gil served as Chief Operating Officer of IBM Research and
the Vice President of AI and Quantum Computing, areas in which he continues to have broad
responsibilities across IBM. Under his leadership, IBM was the first company in the world to
build programmable quantum computers and make them universally available through the
cloud. An advocate of collaborative research models, he co-chairs the MIT-IBM Watson AI Lab,
a pioneering industrial-academic laboratory with a portfolio of more than 50 projects focused
on advancing fundamental AI research to the broad benefit of industry and society.

A passionate advocate of scientific discovery and education, Dr Gil is a member of the
President’s Council of Science and Technology Advisors (PCAST) and is a Trustee of the New
York Hall of Science, which provides schools, families and underserved communities in the
New York City area with exposure to science, technology, engineering and math (STEM).

Dr Gil received his Ph.D. in Electrical Engineering and Computer Science from MIT.

Ushering in a new decade of computing for Smarter cities

Being a “smarter city” is a journey and not an overnight transformation. Government and the urban council must prepare for the changes that will be revolutionary, rather than evolutionary, as they put in place next generation systems which work in entirely new ways. This level of disruption led by AI, high performance computing and even Quantum requires a surrounding compute architecture which can make the entire setup work efficiently, and sustainably. Hear from the Global Head of IBM research how the future of tech is nearer than we think and is being is deployed across the different cities in the world, making it a better place to live.

Senior Computer Scientist, Director of the Urban Center for Computation and Data, USA

Charles Catlett is a Senior Computer Scientist at the U.S. Department of Energy’s Argonne National Laboratory and a Senior Fellow at the University of Chicago’s Mansueto Institute for Urban Innovation. His current research focuses on urban data analytics, urban modeling, and the design and use of sensing and “edge” computing technologies embedded in urban infrastructure. He is the principal investigator of the NSF-funded “Array of Things” (AoT), an experimental urban infrastructure to measure the city’s environment with sensors and embedded (“edge”), remotely programmable artificial intelligence hardware. Operating at over 100 locations in Chicago, AoT is expanding to 200 during summer 2019.

Catlett has served as Argonne’s Chief Information Officer and before joining UChicago and Argonne in 2000, he was Chief Technology Officer at the National Center for Supercomputing Applications at the University of Illinois at Urbana-Champaign. From NCSA’s founding in 1985 he participated in the development of NSFNET, one of several early national networks that evolved into what we now experience as the Internet. During the exponential growth of the web following the release of NCSA’s Mosaic web browser, his team developed and supported NCSA’s scalable web server infrastructure. He is the founding director of the Urban Center for Computation and Data at the University of Chicago and is a Computer Engineering graduate of the University of Illinois at Urbana-Champaign.

Understanding Cities through HPC Supporting Embedded AI, Modeling, and Predictive Analytics

Urbanization is one of the great challenges and opportunities of this century, inextricably tied to global challenges ranging from climate change to sustainable use of energy and natural resources, and from personal health and safety to accelerating innovation and education. There is a growing science community—spanning nearly every discipline—pursuing research related to these challenges. For many urban questions, there is a need for measurements with greater spatial and temporal resolution than is currently available for understanding air quality, microclimate, vibration, noise, and other factors. Concurrently, many factors—from the flow of people through a public space to the impact of at-grade rail crossings on emergency response—require more sophisticated “measurements” involving embedded, or “edge” computing within the urban infrastructure. Ultimately, these and other data sources are also critical to effectively modeling individual urban processes and systems as well as their interactions—requiring high-performance computation. With exascale systems, such coupled models can provide exploratory capabilities through both high-fidelity models as well as ensembles, such as identifying buildings that are particularly vulnerable to extreme heat waves, or transportation investments and policies that are most likely to address congestion and associated heat and emissions challenges. Catlett will discuss initiatives and projects that provide a glimpse into how high-performance cyberinfrastructure will be increasingly central to capturing the opportunities, and addressing the challenges, of urbanization and climate change.

 

Plenary Speakers

Director of the Australian National Computational Infrastructure (NCI), Conjointly Professor of Computational Nanomaterials Science and Technology, Australian National University

Sean Smith is Director of the Australian National Computational Infrastructure (NCI) and conjointly Professor of computational nanomaterials science and technology at the Australian National University. He has extensive theoretical and computational research experience in chemistry, nanomaterials and nano-bio science and technology. He returned to Australia in 2014 at UNSW Sydney, founding and directing the Integrated Materials Design Centre to drive an integrated program of materials design, discovery and characterization. Prior to this, he directed the US Department of Energy funded Center for Nanophase Materials Sciences (CNMS) at Oak Ridge National Laboratory, one of five major DOE nanoscience research and user facilities in the US, through its 2011-2013 triennial phase. During his earlier career, he joined The University of Queensland as junior faculty in 1993 after post-doctoral research at UC Berkeley (1991-1993) and Universität Göttingen (Humboldt Fellow 1989-1991); became Professor and Director of the Centre for Computational Molecular Science 2002-2011; and built up the computational nanobio science and technology laboratory the Australian Institute for Bioengineering and Nanotechnology (AIBN) at UQ 2006-2011. He worked with colleagues in the ARC Center of Excellence for Functional Nanomaterials 2002-2011 as Program Leader (Computational Nanoscience) and Deputy Director(Internationalisation).

Sean has published over 260 refereed journal papers with more than 17000 citations. In 1998 he was elected Fellow of the RACI. In 2006 he was recipient of a Bessel Research Award of the Alexander von Humboldt Foundation in Germany. In 2012 he was elected Fellow of the American Association for the Advancement of Science (AAAS) and in 2015 he was elected Fellow of the Institution of Chemical Engineers (IChemE). He received his PhD in theoretical chemistry from the University of Canterbury, New Zealand, in 1989.

The impact of Australian HPC in the coming decade

The National Computational Infrastructure (NCI) is Australia’s leading high-performance research data, storage and computing organisation, providing expert integrated services to benefit all domains of science, government and industry. NCI’s mission is to deliver to the Australian research sector the computational resources and enhanced compute and data services it needs to achieve transformational outcomes.

The installation of Gadi – NCI’s cutting-edge supercomputer – represents a generational step up in capacity and capability of Australian HPC. Gadi presents unparalleled flexibility for researchers in Australia by leveraging the power of heterogeneity. Whether harnessing the power of GPUs in machine learning, designing new materials technologies or modelling the Southern Ocean, Gadi will be the HPC infrastructure that Australian researchers rely on for years to come.

In this presentation I will discuss the impact of Australia’s fastest ever supercomputer on the research environment down under, and how this new resource has the potential to impact the future of the HPC ecosystem in Asia.

Professor, Applied Mathematics, Stony Brook University, USA

Yuefan Deng earned his BA (1983) in Physics from Nankai University and his Ph. D. (1989) in Theoretical Physics from Columbia University. He is currently a professor (since 1998) of applied mathematics and the associate director of the Institute of Engineering-Driven Medicine, at Stony Brook University in New York. Prof. Deng’s research covers parallel computing, molecular dynamics, Monte Carlo methods, and biomedical engineering. The latest focus is on the multiscale modeling of platelet activation and aggregation (funded by US NIH) on supercomputers, parallel optimization algorithms, and supercomputer network topologies. He publishes widely in diverse fields of physics, computational mathematics, and biomedical engineering. He has received 13 patents.

Fast and Accurate Multiscale Modeling of Platelets Aided by Machine Learning

Multiscale modeling in biomedical engineering is gaining momentum because of progress in supercomputing, applied mathematics, and quantitative biomedical engineering. For example, scientists in various disciplines have been advancing, slowly but steadily, the simulation of blood including its flowing and the physiological properties of such components as red blood cells, white blood cells, and platelets. Platelet activation and aggregation stimulate blood clotting that results in heart attacks and strokes causing nearly 20 million deaths each year. To reduce such deaths, we must discover new drugs. To discover new drugs, we must understand the mechanism of platelet activation and aggregation. To model platelets’ dynamics involves setting the basic space and time discretization in huge ranges of 5-6 orders of magnitudes, resulting from the relevant fundamental interactions at atomic, to molecular, to cell, to fluid scales. To achieve the desired accuracy at the minimal computational costs, we must select the correct physiological parameters in the force fields as well as the spatial and temporal discretization, by machine learning. We demonstrate our results of speeding up a multiscale platelet aggregation simulation, while maintaining desired accuracies, by orders of magnitude, compared with traditional algorithm that uses the smallest of temporal and spatial scales in order to capture the finest details of the dynamics. We present our analyses of the accuracies and efficiencies of the representative modeling. We will also outline the general methodologies of multiscale modeling of cells at atomic resolutions guided by machine learning.

Industry Plenary Speakers

Senior Vice President of Marketing, Mellanox

Gilad Shainer is an HPC evangelist that focuses on high-performance computing, high-speed interconnects, leading-edge technologies and performance characterizations. He serves as a board member in the OpenPOWER, CCIX, OpenCAPI and UCF organizations, a member of IBTA and contributor to the PCISIG PCI-X and PCIe specifications. Gilad is the Senior Vice President of Marketing in Mellanox as well as the President of UCF Consortium. He holds multiple patents in the field of high-speed networking. He is also a recipient of 2015 R&D100 award for his contribution to the CORE-Direct collective offload technology. Gilad holds an M.Sc. degree and a B.Sc. degree in Electrical Engineering from the Technion Institute of Technology.

In-Network Computing – The Next Generation of Supercomputing

The latest revolution in HPC and AI is In-Network Computing. In-Network Computing is the result of a collaborative industry and academia effort to reach Exascale performance by taking a holistic system-level approach toward fundamental performance improvements. The latest generations of smart interconnects offload both the network functions from the CPU and selected data algorithms. This allowing users to actually run data algorithms on the data while the data is being transferred within the system interconnect, rather than waiting for the data to reach the CPU. This technology is referred to as In-Network Computing. In-Network Computing transforms the data center interconnect into a “distributed CPU”, and “distributed memory,” to overcome performance bottlenecks and enabling faster and more scalable data analysis.

Director, Strategy Planning, Cloud and AI, Futurewei Technologies, USA

Francis brings 20+ years of IT industry experience specialized in server systems design, HPC and cloud scale computing architecture. Before joining Futurewei Technologies as Director of Strategy Planning in 2019, Francis served in Huawei Enerprise USA as Director of Product Management and Hardware Architect. Francis is responsible of driving future direction of cloud and AI compute infrastructure.

Prior to joining Futurewei, Francis had 20+ years experiences with industry leading IT solution providers such as Hewlett-Packard, Sun Microsystems and Oracle.

Empowering Smart Cities with an Intelligent Computing Infrastructure

Smart cities drive the needs for a massive computing infrastructure, from the cloud, to the edge. New generation of broadband fibre and 5G networks connect and aggregate tremendous amount of data. An AI-native, distributed and cloud scale compute infrastructure is becoming essential to process, analyse and make decisions. High performance computing plays a critical role in transforming data centres in order to meet the needs of such a demanding intelligent future. HPC empowers and enables smart cities to function. This talk examines the key success factors and practical considerations in designing such an intelligent computing infrastructure, the backbone of smart cities.

Vice President & General Manager, HPC & AI Solution Segment, Hewlett Packard Enterprise, Singapore

Bill Mannel is Vice President and General Manager of High-Performance Computing (HPC) and Artificial Intelligence (AI), for Hybrid IT, Hewlett Packard Enterprise.

Bill joined HPE in 2014 and is a seasoned veteran of the servers and high performance computing industry. For Bill’s first three years at HPE, the HPC business grew significantly more than the overall market. HPE acquired Silicon Graphics International Corp. (SGI), a pure-play HPC company, closing the integration in November, 2017. At the Supercom-puting 2015 show, Bill was named “Person to Watch in 2016” by HPCWire.

Bill joined HPE from SGI, where he was the VP and GM for Compute and Storage products. Prior to SGI, Bill worked at NASA and the U.S. Air Force on aircraft programs such X-29, F-16, B-1B, and AFTI-111.

Bill holds a Bachelor’s degree in Mechanical from Duke University and an MBA from California State University.

Directions for the Exascale Era

An ocean of data is being generated from billions of connected devices at the Edge, as well as larger and larger HPC systems driving simulation and modeling. The world clearly needs new approaches to extract maximum value from all of this data. This sessions lays out the challenges and suggests solutions.

Corporate Vice President and CTO, AMD Datacenter Solutions, USA

Raghu Nambiar is the Corporate Vice President and CTO of Datacenter ecosystems and solution at AMD.

High Performance Computing with AMD

AMD is all about innovation and our mission is to deliver products that help to solve the world’s toughest challenges – in life sciences, earth science, energy, manufacturing, fundamental research, oil and gas, machine intelligence and many more.

 

Creating an inflection point with trailblazing performance and unprecedented scalability for today’s HPC workloads, AMD EPYC processors  and AMD Radeon Instinct mark the next milestone in exascale computing. This session will cover our roadmap, ecosystem partnerships and solutions to address the performance and scalability needs of emerging demands of HPC workloads.

Speakers

University of Amsterdam

Prof.dr.ir. Cees de Laat chairs the System and Network Engineering (SNE) laboratory at the University of Amsterdam. The SNE lab conducts research on leading-edge computer systems of all scales, ranging from global-scale systems and networks to embedded devices. Across these multiple scales our particular interest is on extra-functional properties of systems, such as performance, programmability, productivity, security, trust, sustainability and, last but not least, the societal impact of emerging systems-related technologies. For current activities and projects see: http://delaat.net/

 

ICT to Support the transformation of Science in the Roaring Twenties

The way how science is done is profoundly changing. Machine Learning and Artificial Intelligence are now applied in most of the sciences to process data and understand (or not) the observed phenomena. The recent research directions and results with respect to data and data exchange to feed the AI-ML layer will be addressed in this talk.

Senior Member of the Institute of Electrical and Electronics Engineers, Chief Technologist, Research Networks, CTO Group, Research Labs, Ciena Corporation

Mr. Wilson is responsible for Ciena’s leadership & global interactions with universities and the research community, including national research and education networks. Representing Ciena’s Technology Group, he orchestrates intersections between emerging technologies and Ciena research and development initiatives across the globe. Within Ciena he coordinates a program that extracts new technology innovation initiatives and helps prove viability through testbed trials and high performance demonstrations.
Prior to his Ciena roles, he was a senior advisor for the CTO at Nortel, and held other advanced technology roles during 13 years with the company, including director of broadband switching and development leader of Optical Ethernet. During his time with Gandalf Corporation, he served as VP corporate communications, Applications Engineer and director of Marketing. Prior to this, he served as lead network architect for the University of Toronto’s global bibliographic research service. At Bell Canada, he helped build Canada’s first packet network. Originally trained in Electrical Engineering at Ryerson University in Toronto Ontario, and is a graduate of the Executive Management school at Stanford University in Palo Alto California.
Mr. Wilson is active on a number of business and volunteer Boards. He is chairman of the Algonquin College Foundation Board of Directors; director, Institute of Electrical and Electronics Engineers, Communications Society NA; Industry representative for Polytechnics Canada and is a member of the Institute of Corporate Directors. In 2018, he was elected to the Board of ENCQOR, a $0.4b Public-Private partnership program to develop Canadian 5G wireless technology leadership. He is a recipient of the Queen Elizabeth II Jubilee medal for his service to Canada.

Geomesh Networks, Connecting Continents and Connecting Schools on a Global Scale

Mr. Wilson will highlight advances in highly scalable optical network transport systems that connect continents and connect research applications overland and under the sea. The focus will be on emerging photonic and software technologies that are speeding deployment of scalable diverse long reach systems.

Leader of Network Development Team of KREONET Center, KISTI

He is a Principal Researcher of Korea Institute of Science and Technology Information (KISTI) which is the national Supercomputing & Advanced Research Network Center. Especially, working for Dept. Advanced KREONET Center. Interesting research areas include in ScienceDMZ & PRP, Network QoS & Network Engineering, Software Defined Network & Future Internet, Cloud Computing & Network Virtualization, Remote Collaboration and so on. And also, He is a chair of APRP(Asia Pacific Research Platform) WG in APAN.

Welcome and Intro to APRP

Networks & Communications Manager, King Abdullah University of Science and Technology (KAUST)

Kevin Sale is a network and security professional with 20 years in the telecommunications, education, finance and energy sectors. He currently serves as the Networks & Communications Manager at King Abdullah University of Science and Technology (KAUST) where he is responsible for strategic development of the network to ensure KAUST remains a global tier I research university. Prior to this he lead the Information Security Governance, Risk, Compliance and Awareness practice also at KAUST.

Bridging the Gap

As the APRP and the GRP continue to expand their reach, the West Asia region including Saudi Arabia continues to be under-served. Too often the great science that is being conducted in the region is stifled by a lack of connectivity options and the difficulty of accessing data. As the model of global collaborative research continues to grow there is a widening gap between those who enjoy abundant connectivity and mature platforms and those who do not. This session will outline what the King Abdullah University of Science and Technology (KAUST) is doing in the region to bridge this gap.

Cloud Team Manager, National Computational Infrastructure Canberra Australia

Andrew has many decades of hands-on technical, diplomatic and logistics experience covering a
wide range of standard and bespoke technologies, languages and applications within Industry,
Government, Academia and Research nationally and internationally.

He chairs the judging panel of the SuperComputing Asia Data Mover Challenge and Co-Chairs the
Cloud Security Alliance: HPC Cloud Security, APAN Program Committee, APAN E-Culture and
APAN Asia Pacific Research Platform working groups.

His current role at the National Computational Infrastructure (NCI) involves working on High Performance
Networks, Computing and Cloud systems. He manages the NCI Cloud Team supporting both the NCI
private and the National Nectar Research Clouds and National Data collections.

Data Mover Challenge Judging, Reflections
and Award Announcement

Research Coordinator
DAIKIN INDUSTRIES, Ltd.

Hiroyuki “HIRO” Itoh has been working for the R&D divisions of DAIKIN, including Boston Technology Office. He experienced the development of acoustics field analysis using boundary element method in 1980’s, advanced adaptive control of air-conditioners w/wo machine learning in early 90’s., and business dynamics for the environmental management by discrete-event simulation in 2000’s.

Currently, his mission is to establish the co-creative business environment with various entities. He is a member of the High-Performance Computing Infrastructure Planning & Promotion Committee of the Ministry of Education, Culture, Sports, Science and Technology, and a Fellow and the President of KANSAI Branch of the Japan Society of Mechanical Engineers.

Managing Director, Pacific Teck Limited

Howard Weiss is an expert on creating solutions for high speed storage utilizing HDDs and SSD/NVMes, job scheduling, interconnect technologies and container based virtualization. Howard has acted as the Vice President of APAC for global IT companies such as storage/Data Direct Networks, interconnect/Voltaire (now Mellanox), data protection BakBone Software (Quest/Dell). Howard was a co-founder of Cofio Software that has been acquired by Hitachi Data Systems (now Hitachi Vantara). Five years ago Howard Weiss incorporated Pacific Teck with a mission to provide cutting edge products to the APAC supercomputer, machine learning and high end enterprise market.

Howard has been a key influencer in the design and supports of many of the Top500 class supercomputers around APAC, including AIST ABCI, currently the 7th most powerful supercomputer in the world. He also helped improve the scheduling capabilities of the Tokyo Institute of Technology Tsubame 3 system. He also works on some of the largest NVMe solid state storage projects in APAC including Australia’s CSIRO 2PB all NVMe high speed storage project.

Howard is a graduate of the University of Michigan. Fluent in Japanese, Howard has been based in Tokyo for close to 30 years.

Advanced High Performance Storage, Virtualization/ Containers and Job Management for Powering Smart Cities for the AI and Machine Learning Era

For HPC to powering intelligent cities a key question is how to fully utilize the ever growing processing power of compute resources due to the increase in the number and power of GPUs, how to remove storage bottlenecks with parallel file systems and NVMe drives, and how to run multiple jobs virtualized in a Containers on a node in a parallel environment. Howard Weiss will speak on experiences in supporting the some of the largest supercomputers in Asia Pacific on implementing some of the fastest storage systems, complex job management and running of jobs in a container system that is designed for parallel workloads. The experience leveraged from these projects can enable improved efficiency of computing resources for environments large and small and will aid HPC to power intelligent cities.

CTO, PBS Works, Altair

Dr. Bill Nitzberg is the CTO of PBS Works at Altair and “acting” community manager for the PBS Pro Open Source Project (www.pbspro.org). With over 25 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 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. When not focused on HPC, Bill tries to improve his running economy for his long-distance running adventures

Technology Executive, SSD Application Engineering,
KIOXIA Corporation

Jeff Ohshima is a technology executive of KIOXIA, formerly Toshiba Memory Corporation, which started operation under its new corporate identity as of October 1st 2019, where he focuses on SSD development and application engineering.

He was previously VP Memory Technology Executive at Toshiba America Electronic Components, currently Kioxia America, Inc., where he focused on flash memory with an emphasis on SSDs. He was also Senior Manager R&D in the Advanced NAND Flash Memory Design Department, responsible for 70 nm, 56 nm, 43 nm, and 32 nm designs. He has been engaged in memory technology for over 30 years, including 20 years on DRAM where he acted as a design lead for application specific memories and technical marketing.

He has served as a Visiting Research Scientist at Stanford University.

Meeting HPC Application Needs With Advanced Storage Technology

Further high performance, low latency, and high density requirements are growing for flash storage. Also, the form factor best suited for a system is increasingly being selected. With these evolutionary technologies, cost performance of computing systems are significantly improving.

A new flash architecture that is able to correspond next generation applications is essential to accommodate the wide range of storage needs of smartphones, mobile computing, and data centers. Kioxia offers leading edge SSD technology with a new storage architecture that is best suited for on-premise, cloud data centers, and especially HPC/supercomputing configuration that achieve high performance and excellent TCO.

Senior Staff of Solution Architect, Mellanox

Ashrut Ambastha is a Sr. Staff Architect at Mellanox responsible for defining network fabric for large scale Infiniband clusters and high-performance datacenter fabric. He is also a member of application engineering team that works on product designs with Mellanox silicon devices. Prior to Mellanox, he worked for Tata Computational Research Labs in India and was involved in architecting the Infiniband backbone for Tata’s HPC system “Eka” that was ranked #4 in Top500 list of SC07. Ashrut’s professional interests includes network topologies, routing algorithms and phy signal Integrity analysis/simulations. He holds an MTech in Electrical Engineering from Indian Institute of Technology-Bombay.

InfiniBand In-Network Computing Technology and Roadmap

The latest generations of smart interconnects offload both the network functions from the CPU and selected data algorithms. This allowing users to actually run data algorithms on the data while the data is being transferred within the system interconnect, rather than waiting for the data to reach the CPU. This technology is referred to as In-Network Computing. The HDR 200G InfiniBand In-Network Computing technology empower the world’s leading supercomputers and is paving the road to Exascale computing. The session will discuss the InfiniBand In-Network Computing technology and performance results, and future plans.

© NATIONAL SUPERCOMPUTING CENTRE (NSCC) SINGAPORE 2019