The compute server can contain up to 32 processors, 32 drives, and 64 PCIe cards. See Understanding the Compute Server.
The hardware architecture enables you to divide these hardware resources into smaller units called PDomains to address the demands of your applications. The PDomain configurations available for SuperCluster M6-32 provide a level of redundancy at the PDomain level. See Understanding PDomains for more information.
You can also select two compute servers for your SuperCluster M6-32 to increase redundancy at the compute server level. See Understanding High Redundancy.