Skip to content Skip to menu

We use cookies

We use cookies to analyze traffic, remember preferences and improve the usability of the website. To give your consent, click on the "I Agree" button.

Settings I agree

You can refuse consent at any time.

  • CS

Department of Information Technology

Complex calculations (Metacenter, LUMI, ...)

Simulations and Calculations Using HPC and Grid Infrastructure

In many fields—from physics and bioinformatics to climatology—the need for fast and accurate computations is continuously growing. This is where HPC (High Performance Computing) comes into play. HPC is a high-performance computing environment that enables large-scale simulations, processing of massive data volumes, and accelerates the development of new scientific discoveries. HPC consists of technologies and architectures designed to solve complex computational tasks that ordinary computers could not complete within a reasonable time. In practice, this involves computing clusters made up of hundreds to thousands of nodes, often connected by a high-speed network and managed by a central job scheduler.

Modern research generates enormous amounts of data and requires advanced simulations that would not be feasible without HPC. Examples of its use include:

    • Climate models for predicting changes in the atmosphere and oceans
    • Molecular interaction simulations in biochemistry and pharmaceuticals
    • Calculations in quantum physics and astrophysics
    • Genomic data processing in medical research
    • Artificial intelligence training and machine learning models

To enable research to make effective use of HPC, it became necessary to connect computing resources across institutions and countries. Therefore, the National Grid Infrastructure (NGI) was established at the national level and is connected to the European Grid Infrastructure (EGI). This allows researchers to run computations across multiple institutions (parallel distribution of computational tasks across Europe) and to share data and tools.

System Type What is it suitable for? Login / Access Documentation
MetaCentrum Grid / HPC: Czech national infrastructure for scientific computing General-purpose computing, grid jobs, simulations, bioinformatics Login (CESNET) Continue here
Karolina EuroHPC: Czech supercomputer Large-scale parallel simulations, AI, machine learning Apply through IT4I Continue here
Barbora HPC: Smaller supercomputer, predecessor of Karolina Medium-sized computations, code optimization Apply through IT4I Continue here
LUMI EuroHPC / pre-exascale: One of the most powerful supercomputers in Europe Very large-scale simulations, parallel computing, climate modeling, genomics Apply through IT4I or Apply through LUMI Access Calls Continue here
NVIDIA DGX H100 GPU server Machine learning, deep neural networks, multi-GPU computing Through MetaCentrum (application required) Continue here
VLQ (Quantum) Quantum computer Quantum algorithms, hybrid quantum-classical computing, quantum computing research Access through IT4I / EuroHPC Continue here

*Storage services such as the LUMI-O S3 object storage (30 PB) can also be used for storing computational data.

Who Can Use Supercomputers and How?

    • Employees of Czech research organizations
    • Excellence-based access = granted based on a research proposal after application approval
    • Open IT4I competitions, regular national calls (3 times per year), EuroHPC calls, institutional and community access etc.
    • For fast access, apply for Fast Track Access (FTA), which allocates resources for up to 4 months and provides a decision within 7 days of application.

What Does a Researcher Need to Know?

Documentation and tutorials for most of the items listed below are available on the providers' websites. To simplify computational workflows, a more user-friendly platform called Lexis2 is being developed. Users can log in, for example, through EUDAT (email verification required), ORCID, and other identity providers. More information is available here.

Required knowledge:

    • Working with the command line
    • Running and writing scripts (e.g. Bash, Python)
    • Remote terminal access using SSH
    • How to submit a job and place it in the queue (understanding basic job parameters)

Training and Tutorials

There are many training courses available for working with supercomputers, offered both by service providers and through European projects. User guides are included in the documentation listed in the table above.

Examples of training: