Each customer has a Tenant and within this tenant, you can have multiple workspaces. Within these workspaces, you can have multiple models which is essentially the application solving a business process (T&Q, FP&A, TPM, SPM, etc.). The cross workspace data link I believe you are talking about is having one model in a workspace A being a source or a target of a model in workspace B. The most prevalent example is a Data Hub (a model) in a workspace supplying data to any spoke model in a different workspace. Why the different workspaces? A couple of reasons, a typical workspace can only be 130 GB (excluding our HyperModels which can go up to 720 GB), so you have a finite amount of space. Also, segregation of duties from a security standpoint. You don't need all of your users being able to get into the Data Hub, only the people who manage it.