The Windows Server team have been posting a series on the new capabilities of failover clustering, each of them giving a brief overview and some useful links to get more details on the features.
Here are the ones they have published so far...
This post focuses on the new Mixed-OS mode for clusters which allows you to start with a Windows Server 2012 R2 failover cluster and then start adding in Windows Server 216 cluster nodes by draining, evicting, upgrading and then adding and resuming each node to eliminate downtime for the physical Hyper-V and Scale-out File Server clusters.
Node Fairness is new to Windows Server 2016, identifying over committed nodes and redistributing the virtual machines to under committed modes with zero downtime.
Stretch Clusters leverage the Storage Replica capability of Windows Server 2016, allowing the synchronous or asynchronous protection replication of data across locations, depending on the available bandwidth and latency.
Clusters can now be created as single-domain, workgroup, multi-domain and workgroup and domain clusters, introducing much more flexibility in the deployment options.
VM resiliency is increased with Windows Server 2016 by adding features to address both compute and storage failures.
"Special" virtual machines can be nominated such as those for Active Directory, so that VM can be up and running before anything else that needs it is started.
This covers the benefits that SMB Multichannel introduce, both to Hyper-converged and Converged cluster configurations, which helps speed up activities such as Virtual Machine Live Migrations.