Kubernetes now supports In-Place Pod Resize, a mechanism that allows CPU and memory adjustments without restarting Pods. This removes a key limitation of vertical scaling and reduces disruption for stateful or latency-sensitive workloads. In this News \& Buzz article, we explain how the feature works, its integration with the Kubelet and container runtimes, and its impact on the Vertical Pod Autoscaler and hybrid autoscaling. Finally, we highlight open research directions, including scheduler coordination, application adaptation, and real-world performance evaluation. More >

Fog-enabled Internet of Things (IoT-Fog) architectures and the Cognitive Internet of Vehicles (Cognitive IOV) are becoming key enablers for next-generation intelligent transport systems. At the same time, Blockchain is used to support integrity, transparency, and trust in distributed vehicular services. In this perspective article, we link these domains by discussing a hybrid Mutable-Immutable Blockchain Architecture in which mutable ledgers at the fog layer are combined with an immutable ledger in the cloud. We first outline an analytical model for Cognitive IOV over IoT-Fog, and then extend it to include Blockchain-related delays, throughput, and tamper resistance in the hybrid architectur... More >

IoT-Fog networks create big challenges in scalability, latency, and security at the edge. This paper gives a simple perspective on how sidechains and artificial intelligence (AI) can help. Sidechains process local transactions near devices, increase throughput, and reduce delay. AI methods, such as Random Forest, detect abnormal traffic, predict workload peaks, and adjust network settings (for example, block time and number of validators). We present clear, small formulas for throughput, latency, checkpoint cost, and security, and we show how these formulas guide design choices (block time $T_b$, confirmations $k$, checkpoint period $\tau$). The goal is an easy model that links design choice... More >

As many IoT systems deploy machine learning models to implement intelligent functions, the reliability and performance assurance of artificial intelligence and the Internet of Things (AIoT) system is becoming a crucial issue. While reliability of AIoT system outputs can be improved by redundancy using multiple input data, the system involves performance and energy overheads that may be unacceptable in real deployment under limited computing resources. To ensure the performance and energy-efficiency of AIoT systems, this paper proposes the queueing models for multi-input AIoT systems in two different architectures, namely the parallel and the shared architectures, and compares them with respe... More >

This perspective article argues that hyperparameters such as learning rate, batch size, numerical precision, and training workers are key determinants of energy scalability in CNN training. These parameters directly influence convergence dynamics, hardware utilization, and training duration, leading to substantially different energy profiles even when comparable accuracy is achieved. Moreover, hyperparameter search itself introduces a significant cumulative energy cost, often exceeding that of the final selected model. By analyzing the interaction between convergence behavior and energy consumption, this work highlights the need to treat energy as an explicit scalability metric and to integr... More >