Operator edge-native clouds can provide a unified compute fabric for multiple networks and use cases
14 December 2021 | Research
Digital transformation is driving enterprise applications and processes to the cloud, and particularly to public clouds. Cloud infrastructure needs to become more distributed and ubiquitous to meet future enterprise needs as demand for cloud-based compute and storage grows. The first generation of distributed clouds is being rolled out in metro data centers to provide enterprises with a greater choice of compute locations, both within and between countries. Such clouds are serving the needs of enterprises that want to process certain types of data locally for regulatory compliance reasons and/or that do not want to incur the cost of backhauling large amounts of data to centralized clouds.
Beyond this first wave of ‘edge clouds’, a highly distributed, edge-native compute fabric will be needed in every conceivable location, not just in a few thousand distributed cloud data centers worldwide. This edge compute fabric is expected to serve the processing needs of a new class of data- and event-driven applications (termed edge-native services) that will support novel user experiences and an unprecedented ability to optimize business processes. Such applications include AI/ML, computer vision and autonomous mobility services, as well as those that support collaboration across multiple contexts and devices, such as multi-player, AR gaming. User/device location matters to these edge-native services, and the demand for location-specific, edge-native cloud compute, storage and connectivity will increase as they become more prevalent.
Operator edge-native clouds can provide a unified compute fabric for multiple networks and use casesDownload (PDF)
Data center network automation for CSPs: key trends, challenges and requirements for the cloud-native networking era
SDN in the age of multi-cloud connectivity: the case for a converged programmable network
Wind River: cloud infrastructure