The Grand Duchy has acquired one of the most powerful supercomputers in Europe. This new machine is capable of processing large volumes of data and carrying out thousands of billions of operations per second. But what is the supercomputer actually used for and in which fields? Decoding.
Karolina, Vega, LUMI and MeluXina are not names of cartoon characters, in fact they are names of supercomputers that all have something in common. They are superpowered and belong to the extended family of the company - EuroHPC. As a reminder, EuroHPC is a European supercomputer project based in the Grand Duchy.
The Luxembourg supercomputer is called MeluXina and its name echoes the famous myth of the mermaid Melusina. It is used for research purposes as well as by companies and is installed at the LuxConnect data centre in Bissen.
Hidden from view, in a large room as clean as an operating theatre and surrounded with walls painted in white, a row of cabinets about ten metres long and over two metres high opens up to visitors. In the bowels of these metal containers, servers, storage disks and processors are stacked high, gently humming 24/7.
You've just had the honour of meeting MeluXina! The characteristics of this computer will make your head spin... With a maximum peak computing power of 18 petaflops (corresponding to 18 million billion computing operations per second), it has an unbelievable computing capacity equivalent to that of over 45,000 personal computers. It ranks among the top 50 HPCs in the world.
The supercomputer can process huge volumes of data and carry out complex high-performance calculations (HPC), i.e. simulations dedicated to scientific challenges in the fields of climatology, engineering, aeronautics, automobile design, energy production, medicine or financial analysis. The computing power of MeluXina allows companies and research organisations (manufacturers, SMEs, start-ups, etc.) to access a wide range of services to implement high value-added HPC projects. At present, more than 200 companies from 15 different economic sectors have already expressed their interest in using MeluXina.
The satellite operator SES, the start-up RSS-Hydro, which specialises in environmental remote sensing, and the lunar exploration company ispace are just some of the businesses that make use of the superpower of this computing colossus.
SES optimises the performance of its satellites
Access to MeluXina's capabilities fosters genuine advances in satellite and space innovation efforts. For the satellite operator SES, the use of the MeluXina supercomputer is particularly interesting for optimising satellite performance and spectrum allocation. In other words, resources such as spectrum and transmission power must be constantly allocated in an optimal manner in order to maximise system performance.
Therefore, to enhance the performance of its satellites, SES has to use software systems based on complex algorithms and perform hundreds of tests before these new satellite systems become operational. These tests require high levels of computing power. Thanks to MeluXina, SES can test various multi-orbit scenarios in real time and easily overcome the limitations of conventional computing capabilities. As such, the computing power of the supercomputer makes it possible to generate the optimisation scenarios that enable SES to provide the best possible service to its customers.
MeluXina's services will thus be able to serve millions of users worldwide in every sector imaginable, such as telecommunications, internet services, business and energy, as well as connectivity on board ships and aircraft.
Did you know...
MeluXina is fully powered by green energy and is ranked in the top 15 of the 500 most powerful supercomputers in terms of energy efficiency.
ispace maps the shadows on the Moon
ispace is a lunar exploration company that is currently working on, among other things, a mission to the polar regions of the Moon in search of frozen water. One of the key challenges of the polar environment is shadows. Unlike the Earth, the Moon has virtually no axial tilt. This means that at the poles, instead of having six months of light and six months of darkness, the sun seems to stay right on the horizon all year round.
Therefore, as the sun is so low in the sky, the shadows are very long and can move in a non-intuitive way. For the majority of rovers and other vehicles on the Moon, staying in the light is extremely important as the sun is their sole source of power. Therefore, in order to plan a successful mission, it is vital to be able to predict the location of the shadows at any given time.
This is where MeluXina comes into play. ispace uses the mega-computer to calculate the appearance of the lunar horizon when standing at any given point in the polar regions, enabling the company to develop real-time shadow maps of the lunar surface. Previously, it would take ispace about three minutes to map the shadows on the moon, now it takes microseconds. Furthermore, maps of average illuminance already exist, but they are of low resolution. Thanks to MeluXina, this type of map has become much more accurate.
A unique hub
LuxProvide S.A and its 23 employees are responsible for the operation of MeluXina. In order to facilitate access to High Performance Computing (HPC) and to guide and support companies, Luxinnovation, the University of Luxembourg and LuxProvide jointly manage the National Competence Centre for high performance computing. The centre functions as a unique hub for companies by offering customised support and technical expertise to set up HPC projects, especially for companies with limited levels of skill in this area.
RSS-Hydro forecasts flooding
Globally, flooding is one of the most common forms of natural disaster. RSS-Hydro is active in this field and one of its key objectives is to provide scientific flood risk maps to users in order to assess prevention efforts.
In other words, RSS-Hydro uses MeluXina for flood management via remote sensing, i.e. the acquisition of remote information by satellites. With the help of the mega-computer, the company is able to optimise high-resolution simulations. Thanks to MeluXina, the speed and computational performance of its flood simulations has benefited from a tenfold increase, significantly improving the effectiveness and efficiency of its operations.
For its remote sensing calculations, it uses traditional methods and advanced machine-learning models, combined with the latest technological innovations in computer modelling to simulate water-related hazards and their impact. Drones, satellites and, last but not least, MeluXina are part of working tools utilised by RSS-Hydro.