Skip to content
NAISS Logo
Allocations Support Training Software naiss.se

Project Description LUMI Sweden Fall 2026

Here are detailed instructions for what you should include in the project description for the LUMI Sweden Fall 2026 call. If you are re-using a description from an earlier proposal, please highlight changes. We urge you to respect the maximum page limits for all sections.

Overview (0.5 page)

Provide an abstract of the proposed research.

Include information on computations, data origin and planned analysis. If relevant, mention which compute projects other than this that are generating the data (NAISS projects, EuroHPC JU projects, et cetera).

Resource usage, codes and performance (1.5 pages)

Provide in this section a description of which applications you will run on LUMI and how they make efficient use of the requested computational resources. For LUMI-G, it is important to validate that the applications have support for AMD GPUs, and that they run faster on LUMI-G vs LUMI-C. Include quantitative evidence such as scalability studies, GPU vs CPU relative performance measurements, or graphs (e.g., execution time as a function of node count). The rule of thumb is that an application should run at least 4x faster on the same number of compute nodes when you compare LUMI-G and LUMI-C performance, but exceptions can be made.

Specify the computational resources needed, including the estimated number of gpuhours and/or corehours and the systems on which the work will run. These estimates should be supported by calculations.

Example 1: the expected number of jobs, the gpuhours required per job, and the anticipated runtime.

gpuhours = nodes x used_gpus_per_node x wall_time x number_jobs

Example 2: the expected number of jobs, the corehours required per job, and the anticipated runtime.

corehours = nodes x used_cores_per_node x wall_time x number_jobs

If the requested resources differ significantly from previous allocations or from previously used resources, explain and justify the changes, including any requirements for specialised hardware such as large memory nodes or GPUs. If the project is not a direct continuation of an earlier one, or if you have not previously been granted resources for similar work, include relevant records of past usage. The proposal must also include a detailed description of all required software, indicating whether it must be installed on the system or supplied by the users. Additionally, clarify any licensing requirements and confirm whether the requested resources support the necessary licenses.

Data management plan (1.5 pages)

This section should in particular describe how storage and backup of data and metadata are safeguarded during the research process, how long-term storage is safeguarded, and by whom. Users are encouraged to organise their data systematically and maintain clear documentation of file contents.

Include at least the following:

  1. Please describe the type of data to be managed. Describe the type/types of data to be used in the project, including how data are to be generated/collected. How will data be captured (e.g.: questionnaires, interviews, observations, measurements, recordings etc.). If necessary, has the project passed an ethical review? Can the data be considered sensitive personal data? Include quantitative information when available, such as estimated initial data volume, expected growth throughout the project, and approximate number of files, to justify requested NAISS storage. The parallel file systems on LUMI cannot handle large amounts of small files, please review the limitations listed in the LUMI Documentation and address how you intend to work around them (e.g. by using container images) if you work with many small files.
  2. Please briefly describe data flows, including processing, and tools which may be used to manage data. E.g. How will you organise data that is collected/reused? Will data be organised in databases? Will a specific software or tool be required to use the material? Describe how data will be handled within the NAISS environment, including how ephemeral (scratch) storage versus project storage will be used. Provide a brief motivation for how the requested storage will be efficiently used based on workflow, file size, production rates, and processing steps. If relevant, explain how data will be transferred into and out of the NAISS system.
  3. Please describe the downstream plan for your data. E.g. Where will you transfer the dataset at the end of the allocation? Describe the mechanism by which you will transfer data to external sites for further analysis or archival at the end of the allocation. Give an estimate of the amount of data that will be transferred in each transfer instance. Explain how data will be handled if continued NAISS storage is not granted. If applicable, describe whether data will be made publicly available, for example through a general-purpose repository such as Zenodo.
  4. NAISS strongly recommends keeping a copy of primary research data elsewhere, so briefly describe where primary data will be stored and how backup will be managed. If the project is not a direct continuation of a previous allocation, include a brief record of previous storage usage if relevant to motivate the current request.
  5. Storage is intended only for active data used in analysis, therefore you should explain how outdated or unused data is routinely removed during the project. If storage at a specific NAISS centre is required (e.g. due to hardware, data locality, or workflow constraints), provide justification here or reference the appropriate section.

Scientific challenges (2 pages)

Describe how the proposed project relates to the state of the art of the scientific area, has the potential to increase the knowledge within its scientific area, and how the proposed project uses adequate computational methods to address scientific questions. Include a justification for the need of large-scale data/HPC.

Research group and management (0.5 page)

Describe the complete research group that will be working on the proposed project: name of member, HPC experience, position, and roles. Indicate how the allocated resource will be managed within the group.

References (0.5 page)

Include a list of references for the project description.