Not everything belongs in the public cloud. This is a reality that both customers and major hyperscalers, including Amazon Web Services (AWS), Google Cloud, IBM, Microsoft, and Oracle, are increasingly aware of and planning for. Because of sensitivity to latency, a need for hyperfast response, high security requirements, or bandwidth limitations, a subset of applications will continue to live on-premises or out at the edge. The major hyperscalers have introduced a number of offerings to address these workloads, each taking a different approach. This report describes the current state of these solutions, their differences, and the major scenarios for which enterprises are using them.