An Intelligent Continuum for Intelligent Applications

The compute continuum is considered to be just an aggregation of computational entities encompassing a diverse array of resources, ranging from public clouds to edge devices. To facilitate the development of next-generation intelligent applications, we must transcend this restrictive perspective. In this keynote, we present the notion of an "intelligent meta-continuum" - a purpose-driven, adaptable continuum capable of effortlessly amalgamating distributed resources to execute applications in an optimized and scalable manner. The meta-continuum integrates pivotal technological and architectural advances, such as software-defined infrastructure, workload orchestration, microservices, virtualization, and edge computing. Harnessing these enablers, the intelligent meta-continuum offers a flexible foundation for contemporary distributed applications, allowing them to dynamically adapt to highly variable conditions by optimizing their placement and resource utilization based on factors such as bandwidth, latency, cost, and availability. Innovative programming and organizational models, such as the actor model, play a crucial role in actualizing this vision. These models enable the decomposition of applications into discrete, stateful components that intelligently interact across the meta-continuum. The implementation of such models through technologies like containers and serverless computing allows applications to glide effortlessly across infrastructures. The meta-continuum holds the promise of substantial benefits for next-generation applications in domains such as smart cities, industrial automation, healthcare, and autonomous systems. By seamlessly amalgamating resources from cloud to edge, applications can achieve low latency, extensive distribution, mobility, scalability, and cost-efficiency. To fully unlock its potential, however, we must confront prevailing challenges related to interoperability, security, skill development, and organizational transformation.