Resources

The PLGrid infrastructure provides a wide range of advanced computing and data services that support scientific research and innovation. These resources are designed to meet the diverse needs of researchers , from large-scale simulations and data processing to artificial intelligence (AI), machine learning (ML), and emerging quantum computing applications.

Computing Resources

PLGrid offers access to o a distributed ecosystem of high-performance computing (HPC) clusters located at national supercomputing centers and partner infrastructures, including access to the LUMI EuroHPC system in Finland. The infrastructure includes powerful GPU systems optimized for AI and data-intensive workloads. Users can leverage the latest NVIDIA H100, GH200, and AMD MI250X accelerators for deep learning, model training, and large-scale parallel computations. Traditional CPU-based systems are also available for numerical simulations and scientific modeling across various disciplines. Below, you can find a table containing basic information about available clusters, their hardware configurations, node types, and other technical parameters. More detailed descriptions of each system are available on dedicated subpages, which can be accessed by clicking on the supercomputer names in the table.

* Within the PLGrid infrastructure, access to these systems is limited to selected partitions or a subset of total resources.

Storage Resources

The PLGrid infrastructure provides several categories of storage that support computational workflows across all supercomputing systems. While the general functionality is consistent, the exact configuration, capacity, and performance parameters may differ between systems.

Project space

Project space is available on all PLGrid systems and is typically implemented using the Lustre parallel file system. It is mounted on all compute nodes and intended for storing data associated with active computational projects that require high-throughput access.

Scratch space

Scratch space is provided either at the cluster level or as node-local storage, depending on the system design. It is intended for temporary data generated during computations, and files may be subject to automatic removal based on system policies or inactivity thresholds.

Object storage

Object storage is available in two independent locations, providing redundancy and high availability for large, unstructured datasets. It supports object-based access patterns (e.g., S3), making it suitable for data-intensive workflows, data sharing, and archival use cases that do not require a POSIX file system.

User home directory

Each PLGrid user is assigned a personal home directory, mounted on all systems to which the user has access. It is intended for environment configuration files, source code, and smaller user-specific data.

More detailed and system-specific information about storage configuration, capacity, performance characteristics, and usage policies can be found on the dedicated documentation pages for each individual supercomputer.

Quantum Resources

Researchers can also experiment with quantum computing via the D-Wave cloud platform. This access enables the exploration of quantum annealing techniques, development of hybrid classical–quantum algorithms, and solving of combinatorial optimization problems using real quantum hardware. The service is available through integration with D-Wave’s Leap quantum cloud environment.

---

Last update: December 11, 2025