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

Norway’s National Library is using HPE Cray’s Olivia system and Huawei flash storage to support Norwegian language model training.

A Hacker News discussion questioned whether the setup is powerful enough for full LLM training or mainly suited to fine-tuning.

Commenters debated the project’s goals: building sovereign AI, supporting research, and developing local talent for a small-language ecosystem.

The thread also highlighted the constraints of scarce Norwegian data and limited national compute, which shape how such models can be built.