HPCViz seminar, 26 sept (11.00 – 12.00)

Internet of sports – combining sports and computer science to cool services 

The talk will be divided in two parts: the area of internet of sports and the business opportunities and an example of a service, MySkiLab, for movement analysis of cross country skiing in the field. The service has been developed in close collaboration with the Swedish ski-team and leading researchers in integrative physiology and bio mechanics. The competences needed to realize the services are machine learning, statistics, user experience, big data analytics, client and server programming. The application has now been released in a beta version.

internet-of-sports02-460x272

Dr Christer Norström

Dr Christer Norström is CEO of SICS, Swedish Institute of Computer Science. He has a PhD from KTH in 1997. He has an extensive experience from industry both as engineer and manager with in ABB as well as expert consultant for both Swedish and international companies.

Today his interest is in internet of things and especially its usage within well-being and sport. Christer loves cross country skiing. He is a youth trainer in cross country skiing, author of the waxing book Vallaguiden and the entrepreneur behind the movement analysis tool MySkiLab.

HPCViz seminar, Sept 25 (11.00 – 12.00)

Spark Streaming: Fault-tolerant Streaming Computation at Scale

Matei Zaharia, UC Berkeley

September 25, 2013, 11:00 – 12:00 in the Visualization Studio

Many “big data” applications need to act on data arriving in real time. Running these applications at ever-larger scales requires parallel execution platforms that automatically handle faults and stragglers. Unfortunately, current distributed stream processing models provide fault recovery in an expensive manner, requiring hot replication or long recovery times, and do not handle stragglers. We propose a new processing model, discretized streams (D-streams), that overcomes these challenges. D-streams support a parallel recovery mechanism that improves efficiency over the traditional replication and upstream backup schemes in streaming databases, and also handles stragglers. We show that D-streams can support a rich set of streaming operators while attaining high per-node throughput similar to single-node systems, linear scaling to 100 nodes, sub-second latency, and sub-second fault recovery. Finally, the D-stream model can seamlessly be composed with batch and interactive query models for clusters (e.g. MapReduce), enabling rich applications that combine these modes. We have implemented D-streams in Spark Streaming, an extension to the Spark cluster computing framework.

Logistic Regression: This is an iterative machine learning algorithm that seeks to find the best hyperplane that separates two sets of points in a multi-dimensional feature space. It can be used to classify messages into spam vs non-spam, for example.

Matei Zaharia

 

 

 

 

 

 

 

Matei Zaharia finishing his PhD at UC Berkeley, where he worked with Scott Shenker and Ion Stoica on topics in large-scale data processing and cloud computing. After Berkeley, he will be starting an assistant professor position at MIT. During his PhD, Matei has also been an active open source contributor, becoming a committer on the Apache Hadoop project and starting the Mesos and Spark projects.

HPCViz seminar, December 5, 9.00 – 11.00

Next HPCViz seminar will be held Wednesday December 5: starting at 9:00 in the VIC studio. We are happy to announce the following speakers: Rossen Apostolov and Ali Gholami. See below for more information.

See you there!


 Rossen Apostolov

Title: Ensemble Computing and Markov State Models in Molecular Dynamics Simulations

Abstract:
Molecular Dynamics (MD) simulations are extensively used for atomistic  level studies of structural changes in molecules, inter-molecular interactions, phase transitions etc. A single MD simulation follows the time evolution of a stochastic system and is in practice unable to fully explore all possible states of the system, as needed for achieving statistical significance. The system can reach meta-stable states that  it cannot escape or some events of interest may occur on timescales far beyond what is achievable with current methodology. In addition, most of the supercomputers nowadays have capabilities far beyond the hard-scaling limits of many bio-molecular systems of interest.

Being inherently embarrassingly parallel in nature, ensemble computing presents opportunities for improving the sampling of system states by staging thousands of system replicas to run in parallel. However extracting meaningful information from the vast amount of this trajectory data is not an easy task. Markov State Models (MSM) generate  kinetic
models based on collected large-scale simulation data. They can  help greatly in understanding the evolution of investigated system while allowing to reach timescales larger than those of the individual simulation runs.


 

Ali Gholami

Title: Security of the Biobank Cloud

Abstract:

Recent improvements in the cost and throughput of DNA sequencing machines have caused a mismatch between the increasing rate at which they can generate genome data and the ability of existing tools and computational infrastructure to both store and analyze this data. Biobank Cloud as a new project in context of 7FP aims to address shortage of storage and computational resources through cloud computing service models. However, EU directives on data protection hinder users and organizations to exploit capabilities of cloud computing. In this seminar, we present the Biobank Cloud security concerns and factors that need to be addressed at both interface and internal level. In addition, we discuss authentication, authorization and  auditing mechanisms that are proven in large scale distributed systems.

Exhibition of projects from KID/PEL (Co-operative IT-Design course) at KTH

You’re all welcome to see the students excellent work in this year’s KID/PEL course. The theme this year has been Networking without Borders, which is a starting point for exploration into both technical and social domains. And that is something the students really have done!
There will be six projects presenting (tentative names and descriptions below) in the form of an exhibition and we transform the studio into a fair.
• Patient-relative relations
The next of kins to geriatric patients can get the relevant updates on how their relatives are doing
• Chatroulette for unemployed
A service to simplify the process of linking companies and job seekers by using video 
• Campunize
A social linking and events tool for students
• Neibling
A service for shared services between neighbours
• Atingi
A service for people who share the same dreams and want to make them happen
• Recipes and traditions
A tool for storing, sharing and remembering traditions within an extended family
Where: the Visualization Studio (KTH, Lindstedtsv 5, floor 4, down the corridor)
When: Dec 4, 4 PM – 6 PM
There will be some snacks and drinks.

Welcome!

Seminar with Dr Martin Kronbichler

Welcome to an HPCViz seminar!

Time: Wednesday October 24, at 10:15
Place: The VIC studio, Lindstedtsvägen 5

We are happy to have Dr. Martin Kronbichler from Uppsala University visit
KTH: http://www.it.uu.se/katalog/martinkr

Title: Parallel and efficient finite element solvers for flow problems

Abstract: Simulation tools for flows coupled to scalar transport are
presented. The considered applications are convection in the earth’s
mantle and two-phase flow. The solvers use finite elements on adaptive
meshes that are fully distributed between the computational nodes. I will
present scaling experiments with up to 16,384 processors that show good
scaling in all aspects of simulation: mesh adaptation, assembly, and
solution of linear systems. Moreover, I will talk about improving
efficiency of the implementation by replacing sparse matrices by a
matrix-free scheme. It gives considerably speedup on today’s multi-core
nodes due to reduced memory transfer. The work has been published as a
part of the deal.II finite element library.