News Column

Special Grand Challenge projects are starting

August 4, 2014

ENP Newswire - 04 August 2014

Release date- 01082014 - CSC has approved five projects from the 2014 Special Grand Challenge call.

These projects will be run as Sisu Phase 2 pilot projects during the acceptance tests in August. The planned general availability to all customers of the Sisu second phase is in September. Then Sisu will consist of about 40 000 computing cores and it will reach the petaflop class. The approved projects are:

Planck Full Focal Plane simulation, Hannu Kurki-Suonio, University of Helsinki

The Planck satellite is a mission of the European Space Agency (ESA) to measure the faint echo of the Big Bang. In 2014/15 the Planck collaboration will release its final results, providing powerful constraints on the fundamental parameters of cosmology and high-energy physics and setting the stage for the next generation of Dark Energy experiments.

The basic scientific problem is to determine the cosmological parameters describing the universe as accurately as possible. Important new signatures looked for are possible non-Gaussianity of primordial density perturbations and primordial gravitational waves that can be observed as so-called B-modes of cosmic microwave background (CMB) polarization (these may already have been observed by the BICEP2 experiment, but need to be confirmed by Planck). These signatures are important elements in confirming the cosmological inflation as the origin of structure in the universe and to select among different inflation models

Obtaining precise Planck results will require massive Monte Carlo simulation sets that can only be generated on the largest supercomputers. These simulations are scheduled for summer 2014 and we collaborate with the US NERSC supercomputing center to produce these simulations.

The project will consume 1 000 000 core hours.

AFM nanoindentation of gold nanorods, Kai Nordlund, University of Helsinki

Precise measurement of dimensions and mechanical properties of nanoscale objects like nanoclusters and nanowires is important for emerging technologies and applications. The atomic force microscope (AFM) is often the instrument of choice for these tasks, as it is naturally suited for measurements of mechanical properties. We propose to carry out atomistic molecular dynamics simulations using empirical potentials to simulate the nanoindentation measurement of a gold nanorod on a silicon substrate, using an AFM tip, at the actual scale. This is the only way to reliably compare data from simulation and experiment, and has never been achieved for this type of system before, due to the required system sizes and time scales. The proposed research is part of the EMRP project MechProNO - Traceable measurement of mechanical properties of nano-objects.

The project will consume 2 400 000 core hours.

Gravitational waves from the Early Universe, Kari Rummukauinen, University of Helsinki

Understanding the development of the very early stages of the Big Bang is tightly linked to the understanding of the fundamental nature of matter and interactions. The phase transition associated with the Higgs field happened about a tenth of a nanosecond after the Big Bang, and depending on the physical model used may have given rise to significant gravitational wave background. This radiation provides a unique view of the early Universe, and may be observable in planned space-based gravitational wave detectors such as eLISA. In earlier simulations performed at CSC we have identified the unexpected significance of acoustic waves as the source of gravitational waves. The simulations are restricted by the accessible dynamical range, i.e. the system size. In this Grand Challenge proposal we propose to do a very large volume simulation on a pre-selected set of parameters in order to increase the accuracy of our earlier results and ensure that the dynamical extrapolations are accurate.

The project will consume 720 000 core hours.

Spectral thermal conductivity of micrometer-long carbon Nanotubes, Jani Oksanen, Aalto University

Carbon nanotubes (CNTs) are excellent thermal conductors whose thermal conductivities (TCs) typically increase with the tube length. The details of the scaling of TC and its spectral properties in terms of the tube length are, however, currently not understood. The proposed project applies large-scale molecular dynamics simulations to describe the propagation and scattering of lattice vibrations in CNTs in order to calculate (i) TC of micrometer-long tubes to resolve the uncertainties regarding the scaling of TC, and (ii) the spectral decomposition of TC using the methods recently developed by us [arXiv:1405.3868]. The spectral decomposition provides a clear picture of which wavelengths transfer the heat in long nanotubes and allows for determining the mean free paths of heat carriers at different frequencies. The results of the project are expected to be highly useful for the application of CNTs as thermal interface materials, electrical interconnects and individual CNT devices in future nanoelectronics.

The project will consume 500 000 core hours.

Vlasiator simulation of the bow shock, Rami Vainio, University of Turku

The complex plasma phenomena in the near-Earth space originate from the Sun that emits a variable stream of charged particles carrying the solar electromagnetic field to the interplanetary space. The solar system is a rich plasma laboratory with spatial and temporal scales from milliseconds to decades and centimetres to billions of kilometres. Often the in-situ spacecraft observations can neither distinguish temporal from spatial variations nor provide accurate information concerning feedbacks from other processes. Although the recent multi-spacecraft missions address this challenge, one cannot escape the fact that accurate global modelling remains a key component in modern space research. Present global models are based on the magnetohydrodynamic description of the plasma. However, the newest measuring techniques require directly comparable novel simulations providing a kinetic description of plasma with high-quality ion distribution functions. The newly developed Vlasiator is an answer to this need.

The objective of this project is to comprehensively understand the shock physics within the dayside of the magnetosphere, including the dispersive effects introduced by the Hall term in Ohm's law. While the topic is critical in understanding space weather, high impact is also expected on high-performance computing techniques due to the scale size of the problem.

The project will consume 2 000 000 core hours.

For more stories covering the world of technology, please see HispanicBusiness' Tech Channel

Source: ENP Newswire

Story Tools Facebook Linkedin Twitter RSS Feed Email Alerts & Newsletters