Danish National Supercomputer for Life Sciences

Danish National Supercomputer for Life Sciences
Name	Danish National Supercomputer for Life Sciences
Established	2018
Location	Copenhagen, Denmark
Operators	Danish e-Infrastructure Cooperation (DeIC)

Danish National Supercomputer for Life Sciences is a national high-performance computing installation serving biomedical research, genomics, and translational studies. It supports projects across institutions such as the University of Copenhagen, Aarhus University, Technical University of Denmark, and Statens Serum Institut, enabling collaborations with international partners including EMBL, EBI, CERN, and EuroHPC. The facility integrates infrastructure, software, and data stewardship to accelerate workflows in precision medicine, structural biology, and population genomics.

Overview and Purpose

The system was created to provide compute and storage resources for life-science initiatives involving University of Copenhagen, Aarhus University, Technical University of Denmark, Statens Serum Institut, Aalborg University, and Novo Nordisk Foundation-funded projects. It supports large-scale pipelines used by European Molecular Biology Laboratory partners, European Bioinformatics Institute, European Genome-phenome Archive, Nordic EMBL Partnership, and clinical networks tied to Rigshospitalet and Odense University Hospital. The platform enables integration with international consortia such as Human Cell Atlas, 1000 Genomes Project, Genome Aggregation Database, European Nucleotide Archive, and GA4GH to facilitate reproducible analyses in phylogenomics, proteomics, and metabolomics.

History and Development

Initiated following strategic recommendations from Danish e-Infrastructure Cooperation and funding announcements by the Ministry of Higher Education and Science (Denmark), the project built on national investments similar to those supporting Danish National Supercomputer for Life Sciences partners in earlier initiatives with DeiC HPC, Nordic Data Grid Facility, and PRACE. Development milestones involved procurement processes shared with EuroHPC JU frameworks, vendor engagements with manufacturers like Hewlett Packard Enterprise, Dell Technologies, and NVIDIA Corporation, and software collaborations referencing projects led by European Grid Infrastructure and ELIXIR. Academic steering drew on expertise from principal investigators at University of Copenhagen Faculty of Health and Medical Sciences, Aarhus Universitet Hospital, and members of the Danish Council for Independent Research.

Architecture and Technical Specifications

The architecture combines compute nodes with CPU families from Intel, AMD, and accelerators from NVIDIA (Tensor Core) to support deep learning frameworks such as TensorFlow and PyTorch. Storage subsystems implement parallel filesystems similar to Lustre or GPFS and object stores compatible with Amazon S3 APIs to host datasets like GenBank, RefSeq, and UniProt. Interconnects use high-throughput fabrics driven by InfiniBand or Omni-Path technologies, and orchestration layers leverage Slurm Workload Manager, Kubernetes, and workflow engines inspired by Nextflow, Snakemake, and Cromwell. Security and compliance are aligned with standards promoted by General Data Protection Regulation stakeholders and clinical data frameworks used by European Medicines Agency collaborations.

Research Applications and Use Cases

Common applications include whole-genome sequencing and variant calling pipelines used by groups collaborating with Broad Institute and Wellcome Sanger Institute; cryo-electron microscopy processing akin to workflows at Max Planck Institute for Biophysical Chemistry; molecular dynamics simulations comparable to studies at Lawrence Livermore National Laboratory and Argonne National Laboratory; and single-cell transcriptomics pipelines paralleling analyses at Broad Institute Single Cell Portal. Clinical bioinformatics projects link to trials coordinated by European Clinical Research Infrastructure Network and drug discovery work with partners such as Novo Nordisk and Lundbeck. The platform also supports machine learning research in imaging and diagnostics similar to efforts at Karolinska Institutet and University College London.

Organization, Governance, and Access

Governance is administered through consortia mechanisms resembling models used by DeIC, ELIXIR Denmark, and national research infrastructures such as Danish Centre for Big Data Analytics Integration; advisory boards include representatives from Ministry of Health (Denmark), University of Copenhagen, Aarhus University, DTU Compute, and clinical stakeholders from Rigshospitalet. Access policies follow peer-reviewed allocation procedures modeled on PRACE and EuroHPC calls, with service levels and user support provided through helpdesks inspired by ELIXIR Helpdesk and training coordinated with Nordic EMBL Partnership courses and networks like GOBLET.

Performance, Benchmarks, and Funding

Benchmarking uses community standards from High Performance Linpack and domain-specific suites analogous to those used by Genome Analysis Toolkit workflows and GROMACS benchmarks. Reported performance metrics track FLOPS, I/O throughput, and workload throughput for tools like BWA, Bowtie, STAR, and SPAdes. Funding derives from mixed sources including grants from the Novo Nordisk Foundation, allocations from the Ministry of Higher Education and Science (Denmark), contributions through DeIC, and co-funding mechanisms coordinated with NordForsk and EU Horizon 2020-era programs. Continuous upgrades align with procurement cycles influenced by EuroHPC JU roadmaps and manufacturer roadmaps from Intel Corporation and NVIDIA Corporation.

Category:High-performance computing in Denmark