Starcloud, an AI data center startup backed by Nvidia, said it has trained the first large language model in space.
“This is a significant step toward moving almost all computing off Earth to reduce the burden on our energy supplies and take advantage of abundant solar energy in space,” Starcloud Chief Engineer and co-founder Adi Oltean said on X.
AI models’ intensive computations consume plenty of energy, which has raised environmental concerns.
Oltean said not only did they train an LLM – nanoGPT from OpenAI co-founder Andrej Karpathy – they also ran inference on it, meaning they prompted it to generate content.’