Global load balancing is a service that directs inbound connections to a nominal application/service to actual instances of the application/service, typically at geographically dispersed locations and in accordance with a configured traffic-distribution policy . The global-load-balncing service iteself is typically deployed in a globally avaiable manner.
The distinction between “global” and “application” load-balancing is not a precisely defined one. Depending on your specific use-cases and design objectives, you may find yourself more interested in the global nature of the endpoints to which a GLB directs traffic or the global nature of the GLB’s availability. (Both are elements that typically differentiate global load-balancing services from “application load balancing services.”)