The 16th HPC Connection Workshop
@ ISC17

The HPC Connection Workshop is an international High Performance Computing summit organized by the Asia Supercomputer Community and Inspur Group. This event takes place three times a year: during the ASC in China, the ISC in Germany, and the SC in America. We are committed to the exploration of application innovation and current trends in emerging fields of science and engineering. During the workshop, top researchers and leading professionals from around the world gather together to discuss disruptive technologies and latest development in supercomputing.
Previous workshops:

Machine Intelligence and Supercomputing II

June 19, 14:30-18:30, 2017
Meeting Room Analog 1, Forum Building, ISC Messe Frankfurt, Germany


Deputy Director, Lawrence Berkeley National Laboratory
Chairman, ISC 2018
Supercomputing and Superintelligence
In recent years the idea of emerging superintelligence has been discussed widely by popular media, and many experts voiced grave warnings about its possible consequences. This talk will use an analysis of progress in supercomputer performance to examine the gap between current technology and reaching the capabilities of the human brain. In spite of good progress in high performance computing (HPC) and techniques such as machine learning, this gap is still very large. I will then explore two related topics through a discussion of recent examples: what can we learn from the brain and apply to HPC, e.g., through recent efforts in neuromorphic computing? And how much progress have we made in modeling brain function? The talk will be concluded with my perspective on the true dangers of superintelligence, and on our ability to ever build self-aware or sentient computers.
Professor, Sun Yat-sen University
Director, National Supercomputing Center in Guangzhou
TH-Starlight: HPC on Cloud
The HPC applications expanded from traditional fields of global environment change, materials, high energy physics to emerging fields of life science, economy, IoT, and social development. The model of using supercomputers are changed, the way of accessing supercomputers are variable. To fulfill the various requirements, we need to design and implement capable platform for availability and usability of multi-domain applications. This talk will discuss the technologies of HPC on Cloud and share some application cases on Tianhe Starlight platform.
Director, Science & Technology Facilities Council Hartree Centre
Factors Affecting the Adoption of Machine Learning Technologies in Industrial Collaborations
To maximise the impact of a technology, the technology must be converted into an innovation that displaces an existing practice within a community. There is great opportunity for machine learning in this area. I will introduce frameworks for evaluating innovation that can be used to understand the impact and to accelerate adoption within a community. Then discuss case studies of work being carried out within Hartree Centre applying machine learning technologies to healthcare and chemistry and materials discovery.
Deputy Director, Interdisciplinary Centre for Mathematical and Computational Modelling, University of Warsaw
Musings on the Prospects and Limitations of Supercomputing Machines Surpassing or Augmenting Human Intelligence
In this talk I will pose some questions prompted by recent progress achieved in using Deep Learning techniques, and new computing equipment, in areas such as image or speech recognition and ensuing excitement and newly found optimism about promise of Artificial Intelligence. As examples of this trend I will report on activities at the ICM in textual information recognition and medical image recognition with temporal information.
Is AI really closing the gap with human intelligence? Can we imagine machines that will be able to create objects which have no immediate analogue in the natural world? Is it likely a machine will be driven by an urge and desire for self-expression in poetry, fine arts or music? Will (super)-computing machines be ever endowed with free will? How can we construct machines that will be able to perceive the external world through learning, symbols and senses - through an interconnectivity with IoT? How can we ensure special mechanisms for behavioral conditioning, evolutionary modifications in machines and programs? Or is it necessary?
I will also allude to the need to define critical size or minimal critical level of complexity - threshold for an emergence of a "thinking machine" and the "size" of "connectome". The talk will be concluded with a brief report of a three-year InfiniCortex project, a global scale, concurrent, InfiniBand connected supercomputer spanning four continents and many countries - a prototype of a very highly complex computing system with arbitrary size and topology of interconnect ("synthetic connectome").
Coffee Break
Professor, Director of Institute for Massively Parallel Applications and Computing Technology, The George Washington University
Brain Related Processing: An HPC Perspective Tour and Thoughts
Professor, Universitat Politècnica de Catalunya
AI Director, Barcelona Supercomputing Center
AI Portfolio at Barcelona Supercomputing Center
In recent years the idea of merging AI and HPC to produce better results has been thoroughly discussed (e.g. by building chipsets for specifically to enhance AI algorithms). I will present the approach taken at BSC to tackle the issue and the related questions. I will tour along the most representative research questions and applications we are dealing with. The talk will be concluded with my perspective on the dilemma AI for HPC or HPC for AI? 
Associate Professor, Tsinghua University
Deputy Director, National Supercomputing Center in Wuxi
Sunway TaihuLight: Extreme Computing and Big Data
Professor, Director of Center for Research in Extreme Scale Technologies, Indiana University
High Performance Computing for Artificial Intelligence – not just for Neural Nets
Data Analytics and Machine Learning are disciplines of intense interest, almost defining current trends in artificial intelligence with significant accomplishments at hand. The US 2015 Presidential Executive Order, the National Strategic Computing Initiative (NSCI), calls for the unification of High Performance Computing (HPC) and big data applications, including machine learning, on the path to future exascale computing by the early part of the next decade. Even more innovative concepts include neuro-morphic processing that is brain-inspired and takes many forms with the presumption that if we can make the equivalent of a neuron, we can make the equivalence of the human brain and therefore the property of intelligence. With great excitement, such approaches are being pursued in the US and internationally again with interesting accomplishments. It is the speaker’s opinion that none of these strategies will lead to the ultimate achievement of machine intelligence any more than building a chess-playing machine – or, for that matter, a theorem-proving algorithm - will yield a system that can derive general relativity, recognize a metaphor, or prepare a soufflé. It is asserted that intelligence is an algorithm, not the mimicking individual idioms of a lemur’s mental acuity. Fundamentally, machine intelligence requires the ability to understand and manipulate knowledge in response to newfound objective functions unless of course we’re prepared to accept the logical capabilities of a grasshopper. The intelligence algorithm has not been derived; indeed, the metrics of intelligence are ill formed. But prior art does expose many of the functional primitive that machine intelligence will have to be able to perform expeditiously and therefore capabilities required of such a future machine. This presentation will give a hypothesized high level architecture and workflow of an intelligent machine and from these identify the key primitive operations that are likely to be key to ultimately achieving the emergent property of intelligence from non-biological systems.


 Sign up

First Name  
Last Name  
Job Title  
How did you hear about the 16th HPC Connection Workshop?
  * All information required.
        Contact if you have any questions.