Astrolab is sending its FLIP rover to the moon with a high‑performance HPE computer that will run large language models – the kind of AI behind ChatGPT – on the lunar surface. The 480‑kilogram rover is slated to launch as early as November 2026 and will spend about 10 days near the south pole. The AI will analyze camera and lidar data to help the rover navigate and to cut down the amount of data sent back to Earth. HPE’s Spaceborne‑4 computer builds on earlier versions that ran Llama 3.2 on the ISS.
Why it matters
Running LLMs on the moon could let future rovers make real‑time decisions without waiting for Earth‑based commands.
