ACST 2026

Welcome

The program committee is delighted to invite you to the 5th Workshop on Mobility in the Workshop on AI-Cloud Systems and Theories 2026.

Aim and Scope

AI Cloud Systems and Theories (ACST 2026) has the evolution of AI-native cloud architectures and the advancement of foundational theories for AI-cloud integration as the central theme, a new rising and challenging technical environment that mixes the design of AI-optimized cloud infrastructures, intelligent resource management algorithms, and AI-driven service delivery with the complexity of multi-modal data processing, heterogeneous computing resources, and the dynamic demands of AI applications. The emergence of AI-dependent cloud services and the widespread growth of AI-generated data, as well as the ever-increasing adoption of heterogeneous computing chips and cloud-native technologies, are leading to the accumulation of vast amounts of data related to AI model performance, cloud resource utilization, and service quality metrics. In this context, various technical challenges arise, such as efficient heterogeneous computing resource scheduling for AI workloads, algorithm design to correlate AI model training/inference with cloud resource allocation, the online or offline exploitation of large volumes of AI and cloud operational data for performance optimization, investigating the possibility of offloading AI tasks from centralized cloud to edge nodes and vice versa, algorithm design to integrate traffic offloading, content offloading, and AI task offloading as a consequence of AI-cloud data analysis and pattern inference, and the exploitation of multi-connectivity and training-inference integration technologies for QoS, extreme reliability, and low-latency in AI-enabled cloud services. To tackle these challenges, various issues need to be addressed, such as efficient AI-cloud resource management and optimization, heterogeneous computing integration, efficient data processing and storage for AI workloads, users’ and enterprises’ incentives to adopt AI-native cloud solutions, incentives for cloud providers to deploy AI-optimized infrastructures, incentives for service developers to design AI-driven cloud services, efficient and quality-aware AI model deployment and inference, data-driven AI cloud management, collection and management of AI and cloud data in integrated cloud-edge-fog infrastructures, new business models for AI cloud services, security and privacy protection for AI-generated data and cloud-hosted AI models, as well as related operational concerns and regulatory compliance.

ACST 2026 welcomes submissions from both researchers and practitioners from academia and industry that explore challenges and advances in theories, architectures, algorithms, protocols, middleware, and technologies in current and future AI cloud systems. The workshop has a specific focus on all topics that target to advance the integration of AI and cloud computing, with an emphasis on AI-native cloud architectures, intelligent resource scheduling algorithms, data-driven AI cloud management, cloud-edge-fog integration for AI services, AI-optimized storage and computing technologies, and foundational theories for AI cloud systems. In addition, we encourage work-in-progress and position papers that describe highly original ideas, present new research directions, or have the potential to generate insightful and provocative discussions at the workshop, fostering collaboration between academic researchers and industry practitioners to drive innovation in AI cloud systems and theories.