CONFERENCE PROGRAMME

Day 2 | Wednesday, 16 March 2016
08:00 – 09:00	Registration & Welcome Coffee
BRAIN, A.I. & SUPERCOMPUTING  |  Chair: Horst Simon
09:00 – 10:00	Keynote | Could A Brain Ever Have A Mind? Baroness Susan Greenfield, Oxford University, United Kingdom
10:00 – 10:30	From Supercomputing to Superintelligence Roman Yampolskiy, University of Louisville, USA
10:30 – 10:50	Break
EXASCALE & LANGUAGES  |  Chair: Marek Niezgódka
10:50 – 11:35	Keynote | Sequoia to Sierra: The LLNL Strategy Bronis R. de Supinski, Lawrence Livermore National Laboratory, USA
11:35 – 12:05	Exascale Programming Models: Where Are We Now? Barbara Chapman, Stony Brook University, USA
12:05 – 12:35	In the Zone: HPC with Extempore Andrew Sorenson, The Australian National University, Australia
12:35 – 13:35	Lunch
LANGUAGES  |  Chair: Patricia Kovatch
13:35 – 13:55	PCJ – A Java Library for Scalable Heterogenous Parallel Computing Marek Nowicki, Łukasz Górski & Magdalena Ryczkowska, Nicolaus Copernicus University, Poland Piotr Bala, Interdisciplinary Centre for Mathematical and Computational Modelling, Poland
13:55 – 14:35	Panel Discussion | The Brain in Silicon? John Gustafson (Moderator), Baroness Susan Greenfield, Bronis de Supinski, Roman Yampolskiy, Barbara Chapman, Andrew Sorenson & Horst Simon
INTERCONNECTS & TOPOLOGIES  |  Chair: Bronis de Supinski
14:35 – 14:55	A Dynamic Congestion Management System for InfiniBand Networks Fabrice Mizero & Malathi Veeraraghavan, University of Virginia, USA Qian Liu & Robert D. Russell, University of New Hampshire, USA John M. Dennis, National Center for Atmospheric Research, USA
14:55 – 15:15	Strategies for Topology Aware Job Mapping and Scheduling in Multidimensional Torus-based Petascale Systems Jarek Nabrzyski & Kangkang Li, University of Notre Dame, USA Maciej Malawski, AGH University of Science & Technology, Poland
SYSTEM MONITORING  |  Chair: Bronis de Supinski
15:15 – 15:35	Making Large-Scale Systems Observable — Another Inescapable Step Towards Exascale Dmitry Nikitenko, Sergey Zhumatiy & Pavel Shvets, Lomonosov Moscow State University, Russia
15:35 – 15:55	Break
INFINICORTEX  |  Chair: Robert Harrison
15:55 – 16:15	Around The Globe Towards Exascale: InfiniCortex – Past and Present Gabriel Noaje, A*STAR Computational Resource Centre, Singapore
16:15 – 16:45	How Can We Bridge Supercomputing, Next Generation Networks, Big Data and Applications? Krzysztof Kurowski, Poznań Supercomputing and Networking Center, Poland
16:45 – 17:05	The Role of Standards in Supercomputing David Southwell, Obsidian Strategics, Canada
17:05 – 17:35	InfiniCloud 2.0: Distributing High Performance Computing Across Continents Jakub Chrzeszczyk, Andrew Howard, The Australian National University, Australia Andrzej Chrzeszczyk, Jan Kochanowski University, Poland
18:00 – 21:30	Conference Dinner at 1-Altitude
END OF DAY 2

Day 3 | Thursday, 17 March 2016
08:00 – 09:00	Registration & Welcome Coffee
NEW PROCESSOR ARCHITECTURE  |  Chair: Srinivas Aluru
09:00 – 09:45	A Radical Approach to Computation with Real Numbers John Gustafson, A*STAR Computational Resource Centre, Singapore
09:45 – 10:15	A New Processor Architecture for Exascale and Beyond Thomas Sohmers, Rex Computing, USA
10:15 – 10:35	Break
NEW PROCESSOR ARCHITECTURE  |  Chair: John Gustafson
10:35 – 11:05	The Last Computing Frontier: Quantum Computing Vern Brownell, D-Wave Systems Inc., Canada
11:05 – 11:35	Tables, Graphs, and Problems John Feo, Pacific Northwest National Laboratory & University of Washington, USA
11:35 – 12:05	TrueNorth, A Low-Power Bio-inspired Neurosynaptic Architecture for Sensory Processing Garrick Orchard, National University of Singapore, Singapore
12:05 – 13:05	Lunch
AUTOMATA PROCESSOR  |  Chair: Barbara Chapman
13:05 – 13:50	Keynote | Automata Processing: A New Paradigm for Computing Srinivas Aluru, Georgia Institute of Technology, USA
13:50 – 14:15	Pushing the Frontiers of Supercomputing with Automata Computing Mircea Stan, University of Virginia, USA
14:15 – 14:30	VASim: An Open Virtual Automata Simulator for Automata Processing Research Jack Wadden and Kevin Skadron, University of Virginia, USA
14:30 – 14:45	Parallel Interval Stabbing on the Automata Processor Indranil Roy & Matt Grimm, Micron Technology, Inc., USA Ankit Srivastava & Srinivas Aluru, Georgia Institute of Technology, USA
14:45 – 15:00	Accelerating Weeder: A DNA Motif Search Tool using the Micron Automata Processor Qiong Wang, National University of Defense Technology, China Mohamed El-Hadedy, University of Illinois Urbana-Champaign, USA Ke Wang & Kevin Skadron, University of Virginia, USA Jack Wadden, University of Virginia, USA
15:00 – 15:20	Break
MANY-CORE   |  Chair: He Bing Sheng
15:20 – 15:50	What Have We Learned About Heterogeneous Supercomputing? Wen-mei Hwu, University of Illinois Urbana-Champaign, USA
15:50 – 16:20	Application Readiness for the Pre-Exascale HPC Phase Stan Posey, NVIDIA Corporation, USA
16:20 – 16:50	Many-Core Approaches to Combinatorial Problems Michael Krajecki, Julien Loiseau, Christophe Jaillet, Francois Alin & Arnaud Renard, University of Reims Champagne-Ardenne, CReSTIC, France
16:50 – 17:10	Elastic Stencil Code for Cloud-based GPU Spot Instances Jun Zhou, Yan Zhang & Weng-Fai Wong, National University of Singapore, Singapore
17:10 – 17:25	Break
AUTOMATA PROCESSOR  |  Chair: Mircea Stan
17:25 – 18:25	Panel Discussion | Special Session to Engage Community Input for an Automata Computing Community Infrastructure Mircea Stan (Moderator), Jack Wadden & Audience
18:25 – 18:30	Closing Remarks Marek Michalewicz, A*STAR Computational Resource Centre
END OF DAY 3

GPU Programming Hands-on: Getting Started with GPUs for HPC & Deep Learning

Time:
9:00am – 5:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning & afternoon tea breaks and lunch

Presenters:
Pradeep Kumar Gupta, Lead HPC & Deep Learning Solutions Architect, NVIDIA, APJ
Gabriel Noaje, Senior Computational Scientist, A*STAR Computational Resource Centre

Abstract:
Learn how to program GPUs. With millions of GPU compute enabled GPUs sold to date, software developers, scientists and researchers are finding broad-ranging uses for GPU computing.

Get hands-on practice with OpenACC. OpenACC allows programmers to use simple compiler directives to identify which areas of code to accelerate, without requiring modification to the underlying code itself.

In this interactive class we will also introduce the rapidly developing technology of Deep Learning accelerated by GPUs. Recent advances in Deep Learning have led to a step change in performance in a number of machine perception tasks including visual perception, speech recognition and natural language understanding after decades of slow progress in these areas. We will tour the most popular software frameworks for Deep Learning with goal of helping you decide which framework best suits your application needs as a researcher or developer.

Automata Processor Concepts and Applications

Time:
9:00am – 5:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning & afternoon tea breaks and lunch

Presenter:
Matt Grimm, Automata Processor Applications Engineer, Micron Technology
Indranil Roy, Automata Processor Software Development Architect, Micron Technology
Terry Leslie, Director of Automata Processor Business Development, Micron Technology
Jack Wadden, University of Virginia
Mircea Stan, University of Virginia

Abstract:
This will be a full day workshop, starting in the morning with a brief discussion of concepts and foundational principles. This will incorporate live demonstrations of automata running on hardware. The foundational section will end with a demonstration of Protomata, a protein motif search application. The foundational discussion will be followed by a theoretical discussion of automata processing and mapping various problems into the AP paradigm. Specific applications and research projects will be discussed where the automata processor shows promise. There will also be a discussion of the AP cluster installation at the Center for Automata Processing at the University of Virginia.

OpenPOWER Workshop

Time:
9:00am – 5:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning & afternoon tea breaks and lunch

Presenters:
Norishige (Noly) Morimoto, Vice-President & Chief Technology Officer, IBM Asia Pacific
Ganesan Narayanasamy, OpenPOWER leader in Education and Research, IBM India
H. Peter Hofstee, Research Member, IBM Austin Research Laboratory

Abstract:
OpenPOWER Workshop to provide latest updates on various Industry based applications.

Easily deploy Docker and Openstack with Bright Cluster Manager and Bright Openstack

Time:
9:00am – 5:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning & afternoon tea breaks and lunch

Presenter:
Robert Stober, Director of Systems Solution Architects, Bright Computing

Abstract:
Containerisation and private clouds are gaining popularity in HPC environments. Through this workshop, participants will have the opportunity to learn first hand how to deploy, manage and monitor Docker containers using Bright Cluster Manager. In addition, participants will have hands on experience to deploy an Openstack private cloud “from scratch” using Bright Openstack.

Important notes:
(a) This workshop is limited to 20 participants only.
(b) Participants are required to bring their own laptop.
(c) Prior to the workshop participants should install the Bright Cluster Manager front end client CMGUI for Bright Cluster Manager 7.2 from the here. For help in installing CMGUI, refer to the manual Bright 7.2 Admin Manual excerpt – CMGUI.

Breakthrough Engineering using HPC on Cloud

Time:
9:00am – 1:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning tea break and lunch

Presenter:
Rajesh Chhabra, Vice President APAC – Enterprise Computing, Altair
Srirangam Srirangarajan, Managing Director – Altair Engineering Sdn Bhd.

Abstract:
Engineering is one of the most dynamic sector today. Rapidly changing demands – both in manpower and computing resources – make it also one of the most suitable for leveraging the Cloud. The need to collaborate across the globe with multi-disciplinary experts leveraging the best of the breed engineering applications is an absolute necessity.

These computation intensive engineering applications inturn depend on High Performance Computing (HPC) for quicker turnaround times. Faster turnarounds lead to faster collaboration which in turn would mean faster decision making.

Cloud computing’s unique ability to allow multiple collaborators to work on the same data models improves technical efficiency, data security and reduces potential errors due to miscommunication which in-turn reduces costs and improves overall efficiency.

This workshop is designed for Engineering R&D Heads, Engineering Leads, Product Design specialists and HPC infrastructure providers (technical + management). The workshop will showcase the Cloud solutions for organisations wanting to provide any software as a service and key engineering applications from Altair to enable breakthrough engineering for the modern era using HPC on Cloud.

Performance Tuning on Intel Multi and Many-Core Architectures

Time:
9:00am – 5:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning & afternoon tea breaks and lunch

Presenter:
Mukesh Gangadhar, Senior Application Engineer, Intel

Abstract:
Hardware technologies in High Performance Computing are continuously undergoing major changes and rapidly increasing performance capabilities, but the software and the underlying code legacy is often left unchanged or even neglected. This leads to performance gaps and underutilized hardware assets.

In this workshop, gain insights into cutting-edge programming techniques and tools required to achieve the highest performance on Intel® Architecture using C/C++ or Fortran. Also, learn how to write code in order to maximize software performance on current and future Intel® Xeon and Xeon Phi processors (including Broadwell and Knights Landing).

GPU Progress and Developments for HPC Applications

Time:
9:00am – 1:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning tea break and lunch

Presenter:
Stan Posey, HPC Program Manager, NVIDIA USA

Abstract:
Current trends in high performance computing have advanced towards the use of graphics processing units (GPUs) to achieve accelerator speed-ups for numerical operations common among HPC applications, across a range of computational scientific and engineering domains.

In recent years, this trend has led to GPUs becoming mainstream processors for acceleration of industry-leading commercial CAE software from ISVs such as ANSYS, SIMULIA, and other vendors, and for widely used community developed software including OpenFOAM, WRF, COSMO, and several others.

This session will examine:

• Performance characteristics and requirements of HPC algorithms common to application software in the domains of computational mechanics and atmospheric modeling.
• Current state of GPU parallel solvers in production-use from commercial CAE vendors and considerations for optimal applied-use based on modeling and simulation objectives.
• Developments of GPU parallel atmospheric models and a review of the programming strategies deployed including the use of GPU-based libraries developed by NVIDIA and OpenACC – a directives-based programming model that preserves software portability.
• Select case studies will be reviewed that include deployment of GPUs for practice-level HPC and the application benefits they provide – in some cases transforming current modeling and simulation procedures.
• Review and discussion on NVIDIA directions of GPU hardware, application software development environment, and available programming models.

High Performance Computing with Julia

Time:
9:00am – 5:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning & afternoon tea breaks and lunch

Presenter:
Viral Shah, Co-inventor of Julia, Co-founder of Julia Computing

Abstract:
Julia is a high-level, high-performance, open source programming language for technical computing. It provides a sophisticated compiler, distributed parallel execution, numerical accuracy, and an extensive mathematical function library. Julia’s Base library, largely written in Julia itself, also integrates mature, best-of-breed open source C and Fortran libraries for linear algebra, random number generation, signal processing, and string processing. In addition, the Julia developer community is contributing a number of external packages through Julia’s built-in package manager at a rapid pace. IJulia, a collaboration between the Jupyter and Julia communities, provides a powerful browser-based graphical notebook interface to Julia.

This workshop will start with a basic introduction to Julia. It will then cover topics related to key scientific libraries (array manipulation, linear algebra, sparse matrices, FFTs, etc.). The participants will also learn how to write high performance Julia code, and what makes Julia so fast. Towards the end, we will discuss parallelism, multi-threading, and Julia on GPUs. In the second half, participants may translate their existing programs to Julia or work on an exercise to practise their julia skills.

See: http://www.julialang.org/

Speeding up CFD Simulations for Various Industrial Applications

Time:
9:00am – 5:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning & afternoon tea breaks and lunch

Presenter:
Dr. Vincent Chai, Technical Sales Consultant

Abstract:
In this workshop, participants will be able to experience the power and scalability of ANSYS computational fluid dynamics (CFD) solutions when paired with ANSYS HPC tools for various industry problems. In the recent release, ANSYS CFD with HPC was shown to scale from 768 cores to over 129,000 cores at 90% efficiency when solving complex fluid dynamic problems.

This workshop will go through different case studies with various applications for different industrial applications such as data centre cooling, multiphase analysis, turbomachinery studies and more.

Optimizing with HPC for Advanced Mechanical Simulation

Time:
9:00am – 1:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Morning tea break and lunch

Presenter:
Hari Hara, Application Engineer

Abstract:
In this workshop, participants will be able to experience the power of ANSYS HPC, parametric analysis and optimisation tools for mechanical design and simulation.

With ANSYS finite element analysis (FEA) tools, you can simulate real world behaviour of components and sub-systems. The ability to customize, automate and parameterize your simulations to analyse multiple design scenarios, allows you to test different design variations quickly and accurately. High-performance computing (HPC) adds tremendous value to engineering simulation by enabling the creation of large, high-fidelity models that yield accurate and detailed insight into the performance of a proposed design, predicting the actual performance of the product under real-world conditions.

Overcoming Electromagnetics Challenges in Large and Complex Systems with HPC

Time:
2:00pm – 6:00pm

Venue:
Level 4, Matrix Building, Biopolis

Breaks:
Afternoon tea break

Presenter:
Dr. Boyu Zheng, Technical Manager

Abstract:
Modern electronics industry requires innovative approaches to design higher-speed, higher-throughput, better power efficiency and smaller form-factor electronics systems. As we are entering the era of Internet of Things (IoT), wireless communication and networking becomes equally important with the focus on wireless connectivity, antenna design, and electromagnetic interference and compatibility (EMI/EMC) issues. These systems are usually large and complex, therefore a lot of challenges are encountered during the stages of design and optimization.

ANSYS high-fidelity electromagnetic simulation software is ideal for identifying electromagnetic issues early in the design cycle. For example, the software enables engineers to do design space exploration to quickly identify the ideal solution for antenna design and placement. With HPC and ANSYS electromagnetics solution, engineers can effectively increase system performance and product reliability while reducing field failures.

In this workshop, participants will have the opportunity to understand the workflow of applying ANSYS electromagnetics solution to tackle real-world challenges in large and complex systems. There will also be hands-on sessions for participants to test out ANSYS and feel the solving capability with HPC.