A research organization is migrating a genomics analysis workload to Azure. The workload runs on a 200-node Linux HPC cluster that processes petabytes of unstructured sequencing files. The application requires a POSIX-compliant shared file system with very high aggregate throughput (tens of GB/s) and low latency for concurrent parallel reads and writes from all nodes. The team wants a fully managed Azure service rather than building and maintaining their own parallel file system on VMs. Which storage solution should you recommend?