Quiz Question 1

You are deploying a large-scale inference cluster with over 100 nodes. The model weights are updated infrequently. Which storage option offers the most efficient, high-throughput solution by acting as a shared, read-only cache?