Norway’s National Library uses 2PB of Huawei flash storage to train a sovereign Norwegian-language LLM

Norway’s National Library is building a sovereign LLM trained on Norwegian-language books, newspapers, web pages, broadcasts, and other cultural records. The project uses 2PB of Huawei OceanStor Dorado flash storage in its AI data pipeline, drawing on decades of digitization and special agreements that allow training on copyrighted newspaper content.

The effort reflects a growing push for national AI systems that preserve local language, culture, and institutional knowledge rather than relying on English-centric commercial models. It also shows how large public archives are becoming strategic AI assets, with storage infrastructure and data rights now central to AI sovereignty.

Norway's 2 petabytes of Huawei flash storage and LLM training