June 15 - 18, 2021

The Minho Advanced Computing Center (MACC), FUJITSU and ARM have the pleasure to invite you to a MUG: MACC User Group Workshop at the University of Minho on June 15-18, 2021. This event will take place virtually.

Deucalion is a new EuroHPC world-class green supercomputer in Portugal, which will be installed in the MACC, the largest of the four operational advanced computing centres in Portugal - coordinated by the Portuguese Foundation for Science and Technology (FCT). The machine will use ARM-based technology – the Fujitsu A64FX CPU currently used by Fugaku, the fastest supercomputer in the world nowadays.

Registration is now closed.

If you have any questions, please do not hesitate to contact us!

Agenda

UTC+1

June 15, 2021

09:00 - 10:30

About MACC | Rui Oliveira, MACC

An overview of the Arm programming environment | Conrad Hillairet, Arm

An introduction to Arm SVE | Conrad Hillairet, Arm

10:30

Break

11:00 - 12:00

Introducing Deucalion's SuperComputer Architecture | Nuno Leonardo, FUJITSU

Introducing the world's top supercomputer, Fugaku and A64FX | Toshiyuki Shimizu, FUJITSU

14:00 - 16:30

Arm HPC software ecosystem | Conrad Hillairet, Arm

Introduction to Fujitsu Software Stack | Pierre Lagier, FUJITSU

Hands-on

June 16, 2021

09:00 - 10:30

Hints and tips for porting codes to Arm | Conrad Hillairet, Arm

Debugging with Arm DDT | Conrad Hillairet, Arm

10:30

Break

11:00 - 12:00

Profiling and optimization on Arm | Conrad Hillairet, Arm

14:00 - 16:30

Compilation and Optimisation for A64FX processor | Pierre Lagier, FUJITSU

Hands-on

June 17, 2021

09:00 - 10:30

Deep Learning on Fugaku | Masafumi Yamazaki, FUJITSU

Deep Learning Frameworks on Arm hardware | Milos Puzovic, Arm

10:30

Break

11:00 - 12:00

The importance of Big Data for Financial Risk Management | Américo Oliveira, Feedzai

14:00 - 16:30

Present your problem to MACC

June 18, 2021

08:45 - 10:30

High Performance Computing for Science and Innovation Event

About the speakers

Name: Toshiyuki Shimizu

Affiliation: FUJITSU LIMITED

Short Bio: Mr. Toshiyuki Shimizu is Principal Engineer of Future Society & Technology Unit, at Fujitsu Limited. Mr. Shimizu started at Fujitsu Laboratories with the research and development of the AP1000 massively parallel supercomputer system. His primary research interest is in interconnect architecture, most recently culminating in the development of the Tofu interconnect for the K computer. He led the development of Fujitsu’s high-end supercomputer PRIMEHPC series and Supercomputer Fugaku. Mr. Shimizu received his Master of Computer Science degree from Tokyo Institute of Technology in 1988.

Name: Masafumi Yamazaki

Affiliation: Fujitsu Ltd. Fujitsu Research, ICT Computing Lab. Scalable heterogeneous computing Project

Short Bio: Masafumi Yamazaki is a senior researcher at Fujitsu Ltd. He joined Fujitsu's semiconductor division in 1996 and worked on the semiconductor circuit design for products such as high performance custom memory and dynamically reconfigurable processors. Since 2010, he was involved in the development and operation of large-scale distributed systems. In 2015, he was transferred to Fujitsu Laboratories. His research interests are deep learning applications and acceleration of them using parallel computing.

Webpage

Name: Nuno Leonardo

Affiliation: Fujitsu Technology Solutions, Portugal

Short Bio: Datacenter Business Development at Fujitsu. At Fujitsu since 2007, he started as a consultant in datacenter solutions, and since 2015 has assumed responsibility for business development in the processing and integrated systems offerings in Portugal. In addition to his experience in the design and development of datacenter offers, he is also a trainer for Fujitsu partners and employees, for the commercial and pre-sales areas.

Webpage

Name: João Barbosa

Affiliation: Minho Advanced Computing Center

Short Bio: João Barbosa joined the Texas Advanced Computing Center (TACC) Scalable Visualization team in 2011 as a Graduate Research Assistant while pursuing a Dual Ph.D. at the University of Texas at Austin and the Portuguese MAP-i Program in High-Performance Graphics, under the UT Austin Portugal Program.

As a full-time Research Associate at TACC, João has worked in several Scientific Visualization (SciVis) projects ranging from high-level applications such as Gas and Oil to low-level high-performance software packages, in partnership with leading hardware and software companies. His current research focus lies in high-performance real-time in-situ photo-realistic ray tracing for SciVis. He moved back to Portugal last year and settled in Braga to join the Minho Advanced Computing Center (MACC) as a full-time researcher in Scientific Visualization.

Name: Pierre Lagier

Affiliation: Fujitsu Systems Europe

Short Bio: Responsible for scientific and technical management of High Performance Computing international Fujitsu activities in EMEAI region. Previously, as Chief Scientist of Fujitsu labs. Europe (FECIT), Pierre was responsible for the research activities of a multi-disciplinary Japanese-European team in the field of technical and scientific computing. Prior to Fujitsu, Pierre worked for an American Company in Boston, ALLIANT Computer, designing complex system for numerical simulation such as power plants for ENL in Italy, nuclear reactor simulation for EURATOM, simulation of on board real time computer of the Airbus A320. Pierre had his first experience of multi-domains technical computing challenges working for the European Space Agency. He launched a first AI initiative in the late 80’s for optimal satellite life span using AI to control the on board navigation systems. Piere has PhD in Computational Science and Distributed Computing (1986).

Name: Conrad Hillairet

Affiliation: Arm

Short Bio: Conrad Hillairet is a Staff HPC Engineer at Arm. His background is in advanced numerical methods. He joined Arm back in 2017 as a Support Engineer for the Arm Allinea Studio toolchain. He now specialises in performance engineering activities which include helping partners to port their code on Arm based hardware. Conrad holds a M.Eng in Applied Mathematics from INSA Rouen Normandie. He is also Principal Investigator for the MaX Centre of Excellence.

Name: Milos Puzovic

Affiliation: Arm

Short Bio: Milos is Technical Product Owner for Machine Learning and OpenCV at Arm. Prior to Arm he worked at small startup, Mindtrace, as Head of Engineering where his primary role was in designing and development of infrastructure for rapid development and deployment to edge devices and cloud of novel neural models that were trained using semi-supervised approach.

Before Mindtrace Milos led the work on optimizing applications through hardware and software co-design by using machine learning at the Hartree Centre. He was also at Mathworks where he worked on code generation, optimization and verification of high-level models for different types of architecture and Goldman Sachs where he was part of team that implemented automatic trading infrastructure and market making applications for commodities.

Milos has PhD in Computer Science from University of Cambridge where his thesis was on hardware/software interface for dynamic multicore scheduling and BSc with First Class Honors in Joint Mathematics and Computer Science from Imperial College London.

Name: Rui Oliveira

Affiliation: MACC, INESC TEC and UMinho

Short Bio: Rui Oliveira is Associate Professor with habilitation at the Department of Informatics of University of Minho, director of the Minho Advanced Computing Center and appointed as the national expert to the EuroHPC JU, member of the board of INESC TEC and Co-Director of the UT Austin Portugal program. He obtained a PhD degree in Computer Science from the École Polytechnique Fédèrale de Lausanne in 2000. His main research contributions have been in the field of fault-tolerant distributed systems and large scale data management.

He coordinated the H2020 SafeCloud project on secure processing in the Cloud and the FP6 GORDA project on open database replication, two previous national projects on scalable dependable databases, ESCADA and StrongRep, and the U. Minho team at the FP7 CumuloNimbo and LeanBigdata projects. He counts over 100 research papers on large scale and dependable distributed systems, and has served on the programme committee of several first tier conferences. He serves on the Steering Committees of SRDS, Eurosys, Global CENTRA and the Atlantic Interactions Research Centre.