Partner Report – Meet SESAME Net

A distinctive feature of SESAME Net is the diversity of the consortium members who formed the initial network. To introduce who we are, each newsletter will showcase a selection of our partners, and highlight some of their most recent activities.

The Partner Report is a multi-part reportage and this edition will feature Flemish Supercomputing Centre from Belgium and IT4Innovations National Supercomputing Center  from Czech Republic.

IT4Inn-1_A

IT4Innovations National Supercomputing Center – is an important part of e-Infrastructure of the Czech Republic focused on HPC research and services. Its seat is at the VŠB – Technical University of Ostrava. The research programme of IT4Innovations is focused on HPC and contains the following themes: earth science simulations for disaster management, numerical modeling for engineering, physics and chemistry, development of libraries for parallel computing, modeling for nanotechnologies, soft-computing methods, recognition and presentation of multimedia data, safe and reliable architectures, networks and protocols.

The Center operates the most advanced HPC technologies and services and makes them available to Czech and foreign research teams from both academia and industry. The Center operates two supercomputers – Anselm and Salomon. The Salomon supercomputer was put into the operation in July 2015 at which time it was with its theoretical computing power of 2 Petaflops the 40th fastest supercomputer in the world based on the TOP500 list. It is also the largest installation of Intel Xeon Phi co-processors in Europe.

On June 2015 IT4Innovations joined the Intel Parallel Computing Center Program (IPCC). The objective of the IPCC at IT4I is development of highly parallel algorithms and libraries optimized for the latest Intel High Performance Computing (HPC) technologies.

We have recently acquired a new supercomputer named Salomon, which has a large installation of Intel Xeon Phi coprocessors. The Salomon cluster consists of 1008 computational nodes of which 576 are regular compute nodes and 432 are accelerated nodes, each with two Intel Xeon Phi coprocessors. This is what we want to leverage conducting our research,” says Vít Vondrák, primary investigator of the IPCC project.

Apart from development of highly parallel algorithms and libraries the project also focuses on support of HPC community codes. It includes creating interface between IT4I’s in-house code ESPRESO and existing community codes Elmer and OpenFOAM Extend Project. “We want to give special attention to SMEs by enabling their applications but all users of the Elmer and OpenFOAM projects will benefit from the research,” adds Vondrák.

Even though being one of the newest members IT4I was chosen to host the IXPUG Workshop and Tutorials & Intel® PCC Meeting. The week long event will be held on 14th-18th March 2016 in Ostrava, Czech Republic. The program is very rich and includes sessions on optimizations techniques, multi-device and multi-node scalability, programming and runtime models and preparations of applications for Intel Knights Landing (KNL). The last two days of the week are dedicated to tutorials with hands-on labs on many interesting topics such as OpenMP and MPI, hStreams, Knights Corner (KNC) scaling or KNC to KNL migration. More information on the meetings and registration is available here.

The Flemish Supercomputing Centre – was launched in  2007 as a collaboration between the Flemish Government and five Flemish university associations. Many of the VSC employees have a strong technical and scientific background. Our team also collaborates with many research groups at various universities and helps them and their industrial partners with all aspects of infrastructure usage.

Besides a competitive infrastructure, the VSC team also offers training and full assistance with the introduction of High Performance Computing within your company.

In the second half of 2016, a brand new Tier-1 supercomputer of the Flemish Supercomputer Center (VSC) will become operational. The computer will be one of the 150 largest and fastest supercomputers in the world. The Flemish Minister for Science and Innovation has provided the necessary funding of 5.5 million euros. The new Tier-1 will be housed and operated by the University of Leuven.

Via a public tender procedure, the company NEC was selected to build the machine. The new system will have peak performance of more than 600 teraflops, which is delivered by 580 compute nodes with the latest generation of Intel CPU’s connected through a Mellanox EDR Infiniband network. This will result in three times the computing power of the previous Flemish Tier-1. With the increased capabilities of the new Tier-1, scientists will be able to calculate more detailed models: models with finer resolution, more complex interactions within and outside the model, larger parameter spaces can be explored …

The new system will further lift Flemish research to new heights. In addition to academic users, also researchers from industry are welcome to use the new Tier-1 supercomputer. Utilizing the power of High Performance Computing can help businesses to achieve a quicker and cheaper design of new products, more cost-efficient processes and innovative services. The Flemish Supercomputer Center remains dedicated to also offer infrastructure and expertise to industry.

02-11_6-1

Minister Philippe Muyters: “The Flemish Government is committed to investing in research and innovation. Our research institutions are among the best in the world. We can only maintain and strengthen that top position if we keep investing. The supercomputer can play a crucial role in a wide range of fields. It fills me with pride to see that the Flemish Government has played its part in this matter.”

The previous Flemish Tier-1 supercomputer is currently running at Ghent University, but will be decommissioned towards the end of 2016. At a cost of 4.2 million euro, it contains 8,448 Intel cores spread over 528 computing nodes, which are connected by an FDR InfiniBand network. Since it became operational in 2013, this supercomputer has delivered close to 130 million corehours of compute time, which would be equivalent to 14000 years of work on a single core. The machine has been successfully used in scientific fields including Engineering, Weather Modeling, Computational Chemistry, Materials Design, Data Analysis and Bioinformatics. Several industrial users regularly rely on this Tier-1 supercomputer.

Posted in Letter and tagged .

One Comment

  1. Pingback: SESAME Net – HPC for SMEs - SESAME Net

Leave a Reply

Your email address will not be published. Required fields are marked *

CAPTCHA

*