INDUSTRY SPEAKERS

Joseph Curley
Intel, Singapore

Abstract:

Click here for Joseph Curley’s presentation slides.

HPC, which was restricted to Strategic Science for long time has found its ways in to industry applications and established as clear differentiator for the business by introducing Modeling and Simulation as new dimension in finding innovative solutions for industrial challenges. Of late HPC is finding its importance and become an important tool in Business and several analytics based applications under the Big Data title. This has put major demand for affordable, reconfigurable, multipurpose HPC putting pressure on Systems designers ranging from silicon to applications challenging the traditional Practices of innovation in these industries. Looking at these Demands and being responsible player of the computing Industry Intel has been strategically investing in many new areas which would help to find HPC more affordable and solve the larger problems using cost efficient commodity technologies with added supplementary technologies or innovating the core computing itself. This talk is aimed at sharing the Intel’s approach towards solving these challenging demands

Bio Data:
Joseph (Joe) Curley serves Intel® Corporation as senior director of Code Modernization Organization in Intel’s Enterprise and Government Group. His work focuses on programs designed to help ecosystem partners get the maximum benefit now and in the future from modern microprocessors and compute infrastructure.
Joe joined Intel in 2007 to manage planning activities that lead up to the announcement of the Intel® MIC Architecture in May of 2010. Prior to joining Intel, Joe worked at Dell, Inc. leading the Dell Precision Workstation and consumer / small business desktop products, as well as a series of engineering roles. He began his career at graphics pioneer Tseng Labs.

Jang Thye Cheng
Fujitsu, Singapore

Abstract:

Click here for Jang Thye Cheng’s presentation slides.

HPC, Big Data and Machine Intelligence are paving the way for new developments in compute performance. Recent market view of these technologies are not just showing growth in performance, but also in adoption to drive digital transformation, as evident in new use cases today. Fujitsu envisions the convergence of such compute technologies will further drive human centric innovation.

Bio Data:
TBA

Tong Liu
Mellanox Technologies, APJ & China

Abstract:
TBA

Bio Data:
Mr. Tong Liu is The Director of Market Development at Mellanox, APAC; meanwhile he is Asia Director of HPC Advisory Council, a non-profit research organization founded at USA. At Mellanox, he is responsible for overall marketing and solution development at Mellanox Asia-Pacific, including creating solution with partners, build go-to-market strategy, technology promotion and Mellanox brand building in HPC, Datacenter, and Cloud Computing markets. Prior to Mellanox, he worked at HP and Dell headquarters as Senior Software Design Engineer, Consultant, and System Engineer. He has published over 30 publications on HPC and Big Data at international conferences and magazines.

Seetha R. K. Nookala & Vivek Rai
Intel, Singapore

Abstract:
SSF the HPC frame work defined to solve several interoperability and performance variations from a cluster to cluster. Intel with its effort has made many solution partners to join this effort and today offers SSF as the way to go forward to build HPC systems with industry partners. This talk would focus on the SSF frame work and the technologies involved in designing the various building blocks of SSF.

Bio Data:

Seetha R. K. Nookala –
Seetha. Nookala a veteran in HPC community with 30 years of career in HPC coming from working at HPC labs in Dept. of Space ,C-DAC (Indian Govt. Premiere HPC R&D organization).TATA CRL (Delivered 4th Rank system in 2007 )and presently working at Intel in Government and Enterprise , Business Enablement role.
He is awarded prestigious PARAM Award from C-DAC, Pride of India award from Indian Science Congress and JRD TATA Business Innovation award for his HPC community contributions. At Intel Seetha is responsible for Enabling the National HPC Missions and Strategic HPC and AI projects across APAC and Japan.
His career spans as researcher worked on MPP, Tightly coupled, and loosely coupled HPC systems building blocks to overall system Architectures. Worked on Transputer, i 860, Alpha, SPARC and IA Technologies. And ran a 2K node Cloud HPC as service to International Enterprises from Aero, Auto, Seismic, Digital media clients for 5 years. In 2007 at TATA CRL as Chief Architect delivered the First ever IA and IB based loosely coupled system using Public domain HPC stack ‘EKA’ HPC System which Ranked 4th in Top 500 giving boost to the adoption of commodity server based HPC even for Blue sky research, which was earlier dominated by the Custom or proprietary HPC players.

Vivek Rai –
Vivek Kumar Rai: Technical Sales Specialist  with total experience of 11+ Years spread across in the field of HPC Fabric interconnect validation and Storage Area Network. Previous HPC experience as System Test engineer includes  validation  of  host  software  stack  for  Intel’s  40  Gbps  host  network  card, 40 Gbps switch, Fabric Manager and MPI and IPoIB.

Craig Yamasaki
Hewlett Packard Enterprise, Singapore

Abstract:
This session will cover an overview of the HPE & SGI merger – exploring in detail combined synergies, platforms and customer success stories. The session will also cover our view of the industry, solutions and next-generation technology towards Exascale computing.

Bio Data:
Craig Yamasaki is the Director of the Apollo and SGI Product Management and Product Planning team in the HPE HPC Global Business Unit. The organization drives the purpose built HPC server portfolio and HPC software strategy.

Carlos A. Aoki Thomaz
DataDirect Networks, USA

Abstract:
Performance has been for years the focal point of distributed and parallel file system. Uncountable effort has been invested into Lustre file system to bring supercomputing storage to unprecedented performance rates and massive capacity. However, Lustre file system still lacking fine grain control, quality of service and ability to throttle and control I/O per user or jobid basis. Such features are important on environments that requires job completion predictability and minimal I/O performance variability. DDN has developed a Lustre file system feature that implement QoS based on Token Bucket Filtering allowing system administrators to control the RPC rates for users and/or jobids thus becoming more predictable and reliable. While most of the modern HPC software architecture implements highly parallel I/O strategy, a significant number of codes or data flow still relying on serialized I/O events. Another are of development is the improvement of single thread I/O jobs. These two features combined provides a new level of management and job reliability allowing customers to better utilize their storage resources. The presentation discusses these two features in detail, how it’s implemented and possible use cases.

Bio Data:
Technical Product Manager – High Performance File Systems at DDN. With more than 17 years of experience on High Performance and Technical computing, Carlos worked on different areas of High Tech industry including pre sales and Architecture, Consulting & Professional Services and Product management and development. He has been initially active on general High Performance Computing environments including high speed networks and compute clusters supporting projects on different areas of science including operational weather forecasting & climate, computational fluids dynamics and structural analysis. During the last seven years his focus has been on distributed storage architecture and parallel file system implementation supporting customers and projects on different industries. His international experience includes projects in South America, United States of America, Australia, Japan and Asia.

Patrick Wohlschlegel
Allinea, United Kingdom

Abstract:
HPC is facing a period of paradigm change where deep learning is converging with traditional parallel computing methods to demand ever greater scalability and reduced levels of energy consumption. The HPC ecosystem has responded with ground-breaking next generation hardware systems that provide for unprecedented scale potential. But the developers and users of HPC applications face greater pressure than ever before to continuously update their code to achieve such potential, whilst eliminating software inefficiency which hinders progress around energy consumption and time to result. This presentation explores the key challenges facing application developers in dealing with next generation issues such as the diversity of hardware architectures, scalability and resource and energy efficiency. It will discuss the use of development and analysis tools in providing valuable insight when tackling code porting and optimization and in making the most of next generation computing resources.

Bio Data:
Patrick graduated with an MSc from Bordeaux Graduate School of Engineering (ENSEIRB), specializing in parallelism and distributed systems. He started his HPC career with IBM in Paris, working as a pre-sales benchmarker. As he worked to design, build and deploy supercomputers, he discovered and later joined Allinea, now part of ARM. Today, Patrick is based in Warwick (UK) and works as a Senior Engineering Manager, sharing his expertise in HPC tools to users and scientists all around the world.

Marc Hamilton
NVIDIA, Singapore

Abstract:

Click here for Marc Hamilton’s presentation slides.

AI is broadly regarded as the next major technology change in IT. At the core of today’s new style of computing, AI based tools like Microsoft Cortana and Google Now rely on GPU Deep Learning for their super-human accuracy. Earlier changes such as the advent of low cost Linux servers or cloud computing often took years to become mainstream. What’s different today is the speed of adoption. With the availability of GPUs in all major public clouds and open source DL frameworks like Caffe and TensorFlow, GPU Deep Learning has quickly established itself as a new computing model.
– Presentation on AI use case in Autonomous Vehicle (Nutonomy (1st autonomous taxi in the world), or SMART-MIT) – 30 mins
– Use of AI in Intelligent Video Analytics for Smart Nation (CEO/CTO of Xjera) – 30 mins
– Use of AI in Retail (ViSense CTO, visual search) – 30 mins
– Telemedicine and AI for Healthcare (AlemHealth CEO, or NTU School of Bio Science) – 30 mins
– AI in Education (SUTD, SMU) – 30 mins
– SAP Innovation Center (DL in Enterprise software) – 30 mins

Bio Data:
Marc Hamilton is Vice President of Solutions Architecture and Engineering at NVIDIA. He leads a worldwide engineering team responsible for working with NVIDIA’s customers and partners to deliver the world’s best end to end solutions for artificial intelligence, deep learning, professional visualization, and high performance computing. Prior to NVIDIA, Marc worked at HP in the Hyperscale Business Unit and at Sun Microsystems in the HPC and data center groups. Marc holds a BS in Math and Computer Science from UCLA, an MS in Electrical Engineering from USC, and is a graduate of the UCLA Executive Management program.

Yonghua Lin
IBM Research, Singapore

Abstract:
Cognitive system is the platform to provide the deep learning capability for different AI solutions. From IBM point of view, it includes those innovations from system hardware, system software, cloud, and deep learning. In this session, we will use PowerAI (IBM’s cognitive system framework) as an example, to introduce the design principle and how it impacts deep learning training stage and inference stage. We will discuss how some of the recent innovations, such as I/O and accelerator could bring the difference. We will also share how to design the cognitive system with container cloud technologies can achieve productivity, scalability and high availability, for both training and inference. There is an option for a demo (in the IBM booth) for a system called VisionBrain (which is an IBM developed Deep Learning based platform for image/video analysis to provide customized deep learning capability for enterprise customer). The system can be demonstrated together with some slides for differentiation, and link to IBM’s PowerAI (https://www.ibm.com/us-en/marketplace/deep-learning-platform)

Bio Data:
Yonghua Lin is Senior Technical Staff Member and Senior Manager of Cognitive AI System in IBM Research. She is the founder of IBM Supervessel Cloud (www.ptopenlab.com), which provides open services to communities and industry, e.g. accelerator service, big data, cognitive service and DL/ML computing. She has worked on system architecture, cloud computing and cognitive system research for more than 15 years. And now she is focusing on the AI system for industry. She led the research group develop the large scale AI system on cloud and edge, based on the leading technologies like FPGA virtualization, Deep Learning acceleration and Extreme-Low Bit DNN, etc. Besides, she is driving the innovation on deep learning platform for visual comprehension, called VisionBrain. It has been deployed to provide the image recognition service on IBM cloud. In the past more than 10 years, her work covered all kinds of IBM multicore processors,including IBM network processor, IBM Cell processor, PRISM, IBM POWER 6/7/8, etc. She was also the initiator of Wireless Network Cloud with SDR technology in industry, which is called C-RAN today. She led IBM team built up the FIRST optimized cloud for 4G mobile infrastructure, and successfully demonstrated in ITU, Mobile World Congress, etc. She herself has more than 50 patents granted worldwide and publications in top conferences and journals. She is the chair of IEEE Women in Engineering Beijing Section in 2017.

Xinxin Du
Singapore-MIT Alliance for Research and Technology (SMART) , Singapore

Abstract:
In this talk, he will present the recent development of deep learning research in SMART. The main focus will be on the newly developed car detection system by fusing LIDAR and camera vision through deep learning framework. We have achieved state of the art detection accuracy.

Bio Data:
Xinxin Du received his B.Eng degree in 2011 from Nanyang Technological University (NTU) Singapore. From January 2013, he was with the Department of Electrical and Computer Engineering, National University of Singapore (NUS), pursuing a Ph.D degree in the filed of autonomous vehicle. Since July 2016 onwards, he has been with Singapore-MIT Alliance for Research and Technology (SMART) Centre as a SMART Scholar, working on the perception system of autonomous vehicle. His current research interests include: deep learning in perception, image processing and motion control for autonomous vehicle.

Kap Luk Chan
Xjera, Singapore

Abstract:
Big Data Analytics is now a hot topic in academia and industry.  It is fundamental part in building the smart nation. Video Analytics plays an important part as one of the forms of big data analytics. Due to its computational intensive nature, supercomputing technologies including the use of graphics accelerator such as Nvidia GPUs are being extensively use. This is especially so for training the object detectors using machine learning, in particular the deep learning algorithms. In this talk, we share our experience with the audience the application machine learning in developing object detectors for video analytics.

Bio Data:
Dr Kap Luk CHAN obtained his PhD degree in Robot Vision from Imperial College of Science, Technology and Medicine, University of London, London, U.K. in 1991. He was a tenured associate professor in the School of Electrical and Electronic Engineering, Nanyang Technological University (NTU) till 2015. Since then, he works in industry. He is a principal scientist with XJERA Private Limited. His research areas include Image Analysis and Computer Vision, Image and Video Retrieval, Image Semantics and Understanding, Biomedical Signal/Image Analysis for Computer Assisted Clinical Diagnosis. He has served as a reviewer for several international journals such as IEEE Trans. on Multimedia, IEEE Trans. on Pattern Analysis and Machine Intelligence and IEEE Trans on Image Processing, etc. He has published more than 100 papers in international conferences and journals, and in edited books. He is a member of IEEE, IET and PREMIA.

Fanglin Wang
KAI Square, Singapore

Abstract:
The need for a unified video management system has been the talk since the rise of IP surveillance system. With the standardization of video codecs and converging industrial practices, the unification of video management can be achieved with experienced planning and integration. However, for unified video analytics, we observe higher complexity and increasing diversity. Deep learning advancements fuel to more robust and a wider range of video analytics that becomes usable but at the same time introduce more dimensions for unification. In this session, we will introduce our concept and our generic framework to towards unification of video analytic in a large scale fashion.

Bio Data:
Dr. Wang Fanglin is currently the CTO of KAI Square Pte Ltd and an algorithm researcher interested in computer vision, machine learning and deep learning. He is leading to build the open platform which is an ecosystem for video management, video content analytics and system management. Prior to joining KAI Square, he was a research fellow at National University of Singapore working towards to building a brand tracking system over social networks. He had also worked as researchers in China at three international companies including Carestream, Autodesk, and Sharp.

Sara Sandin
Nanyang Technological University, Singapore

Abstract:
Cryo-EM is a powerful technique for structure determination of frozen vitrified macromolecular complexes at near-atomic resolution. Several important developments have contributed to the recent ‘resolution revolution’ in cryo-EM, including direct electron detection, correction of beam-induced motion, as well as improved classification and 3D reconstruction methods. I will discuss biomedical applications, challenges in image analysis of large and heterogeneous datasets and recent advances in GPU-accelerated single particle cryo-EM.

Bio Data:
Sara Sandin (b.1975) is a structural biologist with a long-standing interest in electron microscopy (EM). She studied chemistry (1999-2001) at Stockholm University, Sweden. She obtained a fellowship from the Knowledge Foundation, Sweden to carry out her doctoral research at the Karolinska Institute (2001-2005). Her PhD supervisor was Prof. Skoglund and she worked with cryo-EM and electron tomography. She obtained an EMBO long-term fellowship and a MRC career development fellowship for her post-doctoral research work (2006-2011) at the MRC Laboratory of Molecular Biological (LMB), Cambridge, UK. At LMB she worked in Prof. Rhodes’ group with human telomerase, single particle EM and chromatin. In 2012 she became an Assistant Professor at the School of Biological Sciences, NTU Singapore. There she established a new cryo-EM laboratory for single particle analysis and correlative microscopy of telomeres.

Andrew Underwood
Dell EMC, Singapore

Abstract:
As computing systems become more powerful, the reality of Artificial Intelligence is fast approaching. With this, Dell EMC has developed an Artificial Intelligence system called “ZhuGe” which is powering next-generation researcher across life sciences, manufacturing, and physics. ZhuGe has been designed and deployed in partnership with the Chinese Academy of Science Institute of Automation (CASIA) to provide our customers with a scalable, flexible market-proven Artificial Intelligence architecture, ready to handle compute and data-intensive workloads to drive breakthroughs faster.

Bio Data:
Andrew Underwood leads the Dell EMC High-Performance Computing and Machine Intelligence efforts across Asia-Pacific and Japan. His passion for innovation has driven him to architect innovative technology solutions that have accelerated research and development in world leading companies and research institutes throughout the region.

Ashwin Nanjappa
ViSenze, Singapore

Abstract:
We’ll describe how commodity GPUs are adopted for cloud­based visual search service in very large scale. A hybrid and asynchronized infrastructure leveraging deep convolutional networks and cloud­based GPU servers will be demonstrated. The talk will conclude with video demos of live integrations of ViSenze’s cloud­based visual search API.
As a leading visual search technology provider, ViSenze has indexed more than 200M of images for enterprise API customers and support millions of image search query on daily basis.

Bio Data:
Dr. Ashwin Nanjappa is a Senior Research Engineer at Visenze, an AI startup in Singapore, where he designs and implements deep learning and computer vision algorithm pipelines for ecommerce portals. He received his Ph.D. degree in Computer Science from the National University of Singapore (NUS) in 2012 for his research on 3D Delaunay triangulation algorithms designed for the GPU. From 2013 to 2015, he was a Postdoctoral Research Fellow at the A*STAR BioInformatics Institute (BII), Singapore working on pose estimation computer vision algorithms and systems.

Wei Wang
National University of Singapore

Abstract:
Big data analysis has become a hot topic recently. It is essential to apply machine learning models for some analysis tasks, for instance, sentiment analysis against reviews for analyzing on-line products, and image classification in a food logging application for monitoring user’s daily intake. Extending traditional database systems to support the above analysis is attractive but challenging. First, it is almost impossible to support all machine learning models in a single (database) system; Second, the model training procedure requires expertise with machine learning background, which adds extra burden for database users. I will present a system, called Ra
ki, which integrates Apache SINGA (incubating) with cloud software to provide deep learning services on cloud platforms. Deep learning experts can use Rafi
ki to train different deep learning models and share them with database users; Ra
ki provides Web APIs for database users to use the shared models as a black box in their applications (e.g., in a user-de
fined function). We demonstrate the usage of Ra
ki by training a deep learning model for classifying food images, and applying it to analyze users’ intake history recorded in a database through a food logging application

Bio Data:
Wei Wang is an Assistant Professor in the Department of Computer Science, National University of Singapore (NUS). He got his B.S. from Renmin University of China in 2011 and PhD from NUS in 2017. His research interests include deep learning systems, and applications for multimedia data.

Christopher Muffat
Dathena, Singapore

Abstract:
Information is one of the most important asset in any digitalized industry. In order to protect sensitive data, and to meet regulatory requirements imposed by different jurisdictions, documents in any regulated institutions must be monitored, classified, categorized and labeled internally. In order to facilitate this in the age of ‘big data’, machine learning and algorithmic based solutions are the future of data management and governance.

Bio Data:
Christopher Muffat is the founder and CEO of Dathena Science, a data governance platform developed based on machine learning and artificial intelligence algorithms. Christopher has over ten years of experience in information security risk management, including leading the internal HSBC and SwissLeaks digital forensics investigation for one of the largest data thief that occurred in Europe , as well as conducting research on cyber security risk in safety-critical systems to prevent cyber terrorism for Emirates.
Prior to founding Dathena, Christopher established and managed the Information Risk Management function for Barclays Europe and APAC. He designed information risk regulatory remediation program on identity and access management, information classification, application and infrastructure security, as well as data loss prevention strategy.

Rommel A. Camillo
Huawei Technologies, USA

Abstract:
High performance computing is rapidly finding new uses in many applications and businesses, enabling the creation of disruptive products and services. Huawei, as a global leader in information and communication technologies, brings a broad spectrum of innovative solutions to HPC. This talk is aimed at examining Huawei’s HPC solutions and sharing creative new ways to solve HPC problems.

Bio Data:
Rommel Camillo (RC) is the Senior Manager for Huawei’s Server and Storage Product Line targeted for HPC, Cloud and Big Data Applications. His role at Huawei includes promoting Huawei HPC Products, forming alliances to discover and formulate compelling use cases, planning and strategic product marketing. He worked with a number of HPC companies providing cost effective flash centric storage solutions to improve their overall performance and reduce latency through the use of Non-Volatile Memory Express (NVMe) based Solid State Devices (SSDs). Prior to joining Huawei, Rommel worked at Samsung as Product Planning Manager enabling Samsung’s global SSD business and  Intel as a Platform and Storage Architect working on the first Non-Unified Memory Architecture (NUMA) Xeon CPU as well as the first Solid State Device from Intel. 

Yong Liu
Institute of High Performance Computing, A*STAR, Singapore

Abstract:
Almost most of industries are and will be soon undergoing major transformation brought by artificial intelligence due to increasing computational power, large amount of data and advanced AI algorithms. Based on IHPC’s engagement with industry users from various sectors, this talk will share some examples on how various industry sectors such as advanced manufacturing, healthcare, digital economy and urban solutions will be transformed by artificial intelligence.

Bio Data:
Dr. Liu Yong is working as Capability Group Manager for Artificial Intelligence Group and Scientist in Institute of High Performance Computing (IHPC) at A*STAR, Singapore. He has worked as Principle Investigator to lead several projects on artificial intelligence and machine learning. Dr. Liu Yong holds a PhD degree from National University of Singapore. After PhD study, he has worked as Post-Doc researcher at Royal Swedish Academic of Science. He has received multiple awards and research grants from Singapore government agencies such as EDB and SPRING Singapore. His research areas include artificial intelligence, large scale machine learning, recommender system, cloud computing and computer networks. He is also the co-author of two books.

Sumei Sun
A*STAR Institute for Infocomm Research, Singapore

Abstract:
In this talk, we will present I2R’s cognitive and secure industrial internet of things (IIoT) for providing connectivity to machines, robots, and sensors for data intelligence-assisted, efficient, sustainable, and automated manufacturing. The design challenges are highlighted, and our proposed technologies are presented to overcome these challenges. Finally a few design cases will be shared.

Bio Data:
Dr Sun Sumei is Head of the Communications and Networks Cluster, Institute for Infocomm Research, Agency for Science, Technology, and Research(A*STAR), Singapore. Her current research focus is industrial internet of things (IIoT). Dr Sun published more than two hundred technical papers in prestigious IEEE journals and conferences, and licensed a number of technologies to industry. She is a distinguished lecturer of IEEE Vehicular Technology Society 2014-2018, a Distinguished Visiting Fellow of the Royal Academy of Engineering, UK, in 2014, and a Fellow of the IEEE.

Huynh Phung Huynh
Institute of High Performance Computing, A*STAR, Singapore

Abstract:
Nowadays, increased data processing requirement leads to increasing usage of accelerators (such as GPU, many-core processor or FPGA) to improve performance and throughtput of applications. In this talk, we will share our accelerator computing works at A*Star Institute of High Performance Computing (IHPC) for high performance computing (HPC) and artificial intelligent (AI) needs. It ranges from matrix multiplication, graph processing, streaming processing to object/image identification and recognition using GPU or mobile GPU/FPGA on embedded systems.

Bio Data:
Huynh Phung Huynh received his Ph.D. in Computer Science from the National University of Singapore. His research interest focuses on embedded system, high performance computing (HPC) research such as developing productivity tools for GPU, many cores, FPGA and other accelerators as well as HPC for data mining, natural language processing and machine learning algorithms.

Song Hua Zhang
Infocomm Media Development Authority, Singapore

Abstract:
TBA

Bio Data:
TBA