Software engineer Andrew Rossignol successfully ran generative AI on a 20-year-old PowerBook G4.
The PowerBook G4 had a 1.5GHz PowerPC G4 processor and 1GB of RAM.
By porting the open-source llama2.c project and using the PowerPC vector extension AltiVec, Rossignol achieved inference with Meta's LLM model Llama 2 on the laptop.
Other similar cases of generative AI models running on outdated devices have been reported in the past.