The Norwegian agri-tech company DigiFarm has won two prestigious awards this week at the SC24 conference in Atlanta, one of the world's largest events for supercomputing (HPC).
Each year, the HPC community celebrates exceptional contributions to the field through awards highlighting innovative uses of HPC and artificial intelligence (AI) to address global challenges.
Recognising outstanding achievements and ground-breaking collaborations
DigiFarm received the 2024 HPC Innovation Excellence Award, presented by the renowned HPC analytics group Hyperion Research and their HPC User Forum. This award honours users who have achieved remarkable results with HPC.
The company was also recognised with the Editors' Choice: Best HPC Collaboration (Academia/Government/Industry) award during the HPCwire Readers' and Editors’ Choice Awards 2024. HPCwire, a leading news outlet in the field, bases its awards on a nomination and voting process among readers, as well as selections by its editorial team.
Supporting farmers and agricultural enterprises worldwide
Both awards highlight DigiFarm's pioneering work in transforming global agriculture through HPC and AI. The company leverages the immense computing power of the LUMI supercomputer to process high-resolution satellite data (Sentinel-2 super-resolved, improving resolution from 10 metres to 1 metre). Machine learning models are then trained to detect field boundaries and planted areas based on vegetation and biomass data with precision. On average, DigiFarm processes 560 million hectares of land and 1.6 billion hectares of high-resolution satellite data at 1-metre precision each month.
Accurate and up-to-date data on field boundaries and planted areas are critical for precision agriculture, land digitalisation, and agricultural analytics. Unfortunately, existing public data sources are often outdated and inaccurate. For instance, the Common Land Units (CLU) dataset in the United States, containing 32 million field boundaries, was manually digitised 16 years ago and is no longer reliable. Similarly, the EU’s Land Parcel Identification System (LPIS) is outdated across its 27 regions.
DigiFarm’s technology helps farmers, agricultural enterprises, and governments optimise resource management, reducing costs by 10% while increasing crop yields and efficiency.
Nils Helset, CEO of DigiFarm, expressed gratitude for the awards and emphasised LUMI's value in driving innovation and sustainability:
- LUMI not only supports our work to optimise land use in agriculture but also contributes to sustainable development for society as a whole. For example, our efforts have reduced subsidy control costs for the Lithuanian National Paying Agency (NMA) by 25% annually. NMA oversees over 120,000 farmers and their subsidy applications. This equals approximately €1.5 million in savings and 113 tonnes of annual CO2 emissions (scope 3) avoided by automating subsidy control processes.
Helset added:
- Collaborating with LUMI has significantly accelerated our innovation. We’re particularly grateful for the support provided by Norway’s National Competence Centre for HPC, which ensured a smooth and efficient start with LUMI.
A key role for HPC in AI projects
Roger Kvam from Sigma2, who leads the National Competence Centre for HPC, underlined HPC’s growing importance:
- HPC has become a critical factor in AI projects, delivering significant time and cost savings. Although HPC might seem technically challenging at first glance, it’s more accessible than many realise. Our mission is to help companies like DigiFarm optimise their use of these resources.
DigiFarm exemplifies how national supercomputers, including Norway’s share of LUMI, can drive innovation, enable new products, and support development.
About the National Competence Centre for HPC
Coordinated by Sigma2, the National Competence Centre for HPC offers free support and training to companies exploring HPC and AI opportunities. Through Sigma2, Norway’s national provider of computational resources, academia, public administration, and industry can access supercomputers like LUMI.