Meta sheds more light on how it is evolving Llama 3 training — it relies for now on almost 50,000 Nvidia H100 GPU, but how long before Meta switches to its own AI chip?
CAT
1 min read
Meta reveals how it makes use of almost 50,000 Nvidia H100 GPU to train its Llama 3 LLM.
Originally appeared here:
Meta sheds more light on how it is evolving Llama 3 training — it relies for now on almost 50,000 Nvidia H100 GPU, but how long before Meta switches to its own AI chip?
