The Challenge of Big Data in Science (6th International LSDMA Symposium)

Tuesday 29 Aug 2017, 08:30 → 14:00 Europe/Berlin

Aula, FTU (KIT)

Management of scientific Big Data poses challenges at all stages of the data life cycle – acquisition, ingest, access, replication, preservation, etc. For scientific communities the data exploration – commonly considered as the 4th pillar besides experiment, theory and simulation – is of utmost importance to gain new scientific insights and knowledge.

At this symposium, organized by the cross-program initiative “Large-Scale Data Management and Analysis” (LSDMA) of the German Helmholtz Association, international experts address several aspects of Big Data. It also provides a common platform for discussions and identification of new perspectives.

For further info, please contact lsdma@scc.kit.edu

08:30 Registration

1 Welcome to LSDMA Symposium

Speaker: Prof. Michael Decker (KIT)

Slides 01_2017-08-29_LSDMA-Symposium-Opening,_KIT.pptx

2 The NFFA Europe Information and Data Repository Platform

Speaker: Stefano COZZINI

Slides 02_NFFAatLSDMA-SC.pdf

3 Towards the Fenix Infrastructure

In the context of the Human Brain Project a set of European supercomputing centres have committed themselves to develop and deploy a set of services that will be federated across the involved sites. This effort currently involves five centres from five different countries, namely BSC (Spain), CEA (France), CINECA (Italy), CSCS (Switzerland) and JSC (Germany). The resulting infrastructure will comprise scalable compute resources, data services as well as interactive compute services. While the infrastructure will be made available to several research communities, the Human Brain Project is currently the prioritised driver for the Fenix infrastructure design and implementation.

In this talk we give an overview on the design principles and discuss selected use cases, which the infrastructure need to support. We will provide an overview on the envisage architecture and the strategies for addressing key architectural and technical challenges.

Speaker: Prof. Dirk Pleiter

Slides 03_pleiter-fenix-20170829.pdf

10:20 Coffee Break

4 AI-driven decision automation in physics research and enterprises

A neural network algorithm originally written for physics analyses at CERN is the foundation of many successful projects and products of Blue Yonder, one of the very few AI companies that already deliver huge value to their customers. With its recent focus on building supply chain and pricing products for retailers, strategic management decisions are broken down to tens of millions of automated operational decisions with superhuman quality. The principle and some examples from research and industry are presented.

Speaker: Prof. Michael Feindt

Slides 04_LSDMA2017Feindt.pdf

5 Helix Nebula Science Cloud

The work of Helix Nebula [1] has shown that is it feasible to interoperate in-house IT resources of research organisations, publicly funded e-infrastructures, such as EGI [2] and GEANT [3], with commercial cloud services. Such hybrid clouds are in the interest of the users and funding agencies because they provide greater “freedom and choice” over the type of computing resources to be consumed and the manner in which they can be obtained.

Propelled by the growing IT needs of the Large Hardon Collider, CERN is leading a H2020 Pre-Commercial Procurement activity that brings together a group of 10 of Europe’s leading research organisations to procure innovative IaaS level cloud services for a range scientific disciplines.[4] HNSciCloud is divided in 3 phases: Design, Prototype and Pilot. The successful designs were selected by the end of 2016. In 2017, the project entered the Prototype phase.

This talk will cover the current status of the prototypes and the next steps going into the Pilot phase.

[1] http://www.helix-nebula.eu/

[2] http://www.egi.eu/

[3] http://www.geant.net/

[4] http://www.hnscicloud.eu

Speaker: João FERNANDES

Slides 05_Joao-LSDMA-29Aug2017.pptx.pps

11:50 Lunch Break

6 Accelerating Storage System Research Through a Common Framework

JULEA is a flexible storage framework that contains all the necessary building blocks for storage research. It runs completely in user space, which eases development and debugging. The framework allows offering arbitrary client interfaces to applications; its data and metadata backends can be freely accessed and thus allow rapidly prototyping new approaches.

Speaker: Dr Michael Kuhn

Slides 06_AcceleratingStorageResearch.pdf

7 Research Data and Management in Environmental Sciences - Characteristics, Status and Challenges

Research in environmental sciences is widely 'digitized' and a plethora of approaches and initiatives for environmental (research) data management and infrastructures established. However, still a number of issues hinder seamless data integration for environmental researchers and reproducible research is far from being reality.

This contribution will give an overview on the current status, list some promising approaches as well as current r&d challenges.

Speaker: Prof. Lars Bernard

Slides 07_RDMEnvSciences_LB_TUD_2017_8_komp.pptx

8 Conclusion

Speaker: Prof. Achim Streit (KIT)