With the release of NVIDIA's DGX SuperPOD of A100 GPUs, supercomputers will/are becoming more and more common-place.
What potential deep learning tasks/applications might become more accessible with this large amount of computing? These applications/tasks will require significant computing for either data-extraction, training, etc.
An example of applications requiring large compute would be:
- GPT (or other large NLU/NLP model) Training
- Conducting Neural Architecture Search (NAS)