The potential of superconducting single flux quantum (SFQ) devices in accelerating deep neural networks (DNNs) has garnered significant attention due to their ultra-fast and lowpower switching capabilities.However, existing SFQ-based DNN accelerators face limitations in scaling up to larger-scale instances due to the stringent area constraints and complex architectures.Additionally, another challenge in SFQ-based DNN acceleration lies in bridging the gap between the ultrahigh computing speed offered by SFQ technology and the relatively low memory bandwidth.To address these challenges, we propose JBSA, an SFQ-based bit-serial accelerator for DNN inference acceleration.JBSA leverages bit-serial computing to alleviate area constraints and reduce bandwidth requirements.A bit-serial processing element is designed to implement multiply-accumulate operations using SFQ logic cells.

Due to the surge in mobile apps, online videos, and cloud gaming, central servers struggle with high traffic demands. Content delivery networks (CDNs) mitigate this by distributing content from edge caches, easing backbone network strain. Yet, current allocation algorithms, constrained by practical complexities and billing, often fail to optimally schedule traffic, causing server overload and increased costs. Our solution, CEQC-CDN, addresses these issues by considering DNS load balancing, regional hijacking, quality constraints, and server capacity limits. By incorporating these factors as linear constraints through binary decomposition and variable introduction, CEQC-CDN achieves global optimal traffic scheduling. Employing a greedy policy for monthly optimization, tests demonstrate an 8.57% reduction in CDN traffic costs versus conventional greedy algorithms.

Both Byzantine resilience and communication efficiency have attracted tremendous attention recently for their significance in edge federated learning. However, most existing algorithms may fail when dealing with real-world irregular data that behaves in a heavy-tailed manner. To address this issue, we study the stochastic convex and non-convex optimization problem for federated learning at edge and show how to handle heavy-tailed data while retaining the Byzantine resilience, communication efficiency and the optimal statistical error rates simultaneously. Specifically, we first present a Byzantine-resilient distributed gradient descent algorithm that can handle the heavy-tailed data and meanwhile converge under the standard assumptions. To reduce the communication overhead, we further propose another algorithm that incorporates gradient compression techniques to save communication costs during the learning process. Theoretical analysis shows that our algorithms achieve order-optimal statistical error rate in presence of Byzantine devices. Finally, we conduct extensive experiments on both synthetic and real-world datasets to verify the efficacy of our algorithms.