The eight volume set, LNCS 16381-16388 constitutes the refereed proceedings of the 25th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2025, held in Zhengzhou, China, during October 30 -November 2, 2025.

The 158 full papers, 104 research papers and 48 session papers included in these proceedings were carefully reviewed and selected from 543 submissions. They focus on the following topical sections:

Part I : Parallel and Distributed Architectures; Software Systems and Programming Models.

Part II : Parallel and Distributed Algorithms and Applications.

Part III : Parallel and Distributed Algorithms and Applications; Internet of Things and Cyber-Physical-Social Computing; Performance Modeling and Evaluation.

Part IV : Service Dependability and Security in Distributed and Parallel Systems; Network Architectures and Algorithms.

Part V: Network Architectures and Algorithms.

Part VI: Parallel and Distributed Architectures; Software Systems and Programming Models; Parallel and Distributed Algorithms and Applications; Big Data Management and Analysis; Performance Modeling and Evaluation.

Part VII: Service Dependability and Security in Distributed and Parallel Systems; Network Architectures and Algorithms; Internet of Things and Cyber-Physical-Social Computing.

Part VIII: Intelligent Distributed Computing; Resource Coordination and Joint Optimization in Cloud-Edge-End Systems; Symbiotic AI and Data Ecosystems; Smart Education Powered by Parallel and Distributed Processing; AI for Networks and Networking for AI; Emerging Network Technologies in Computing and Networking Convergence.

.- Parallel and Distributed Algorithms and Applications.
.- Fossil: A Cost-effective and Fault-tolerant Task Placement Scheme for Geo-Distributed Clouds.
.- Towards Privacy-Preserving Collaborative Detection of DDoS with Secure Multi-Party Computation.
.- FedCKD-ALDP: A Dual-Optimization Framework for Non-IID Federated Learning via Clustered Knowledge Distillation and Adaptive Local Differential Privacy.
.- FMQ-ZNS: Enhancing ZNS-Aware Fairness and Performance through Multi-Queue I/O Scheduling.
.- More Resistant and Less Waiting: Parallel Federated Split Learning for Object Detection Missions of UAV Cluster.
.- Multi-Distances Weighted Adaptive Fuzzy K-Nearest Neighbors Algorithm.
.- EAHP: An Efficient Automatic Hybrid Parallelism Approach with Genetic Algorithm.
.- Shadow: Research on Asynchronous DAG Consensus Mechanism Based on Dynamic Privacy Address Selection.
.- GIT: Accelerating Distributed DNN Training via Similar Gradient Filtering.
.- A Lightweight Framework for Energy-Aware Prediction and Scheduling in Heterogeneous HPC Clusters.
.- Hybrid-SAIME: Accelerating Surface and Interbed Multiple Elimination Method via 3-Stage Pipeline Parallel and Load Balancing on CPU-GPU Platforms.
.- HETER-GSD: Accelerating Heterogeneous Edge GNN Inference.
.- SHPTA: Stable Hybrid Parallel Distributed Training Architecture in Dual-Heterogeneous Environments.
.- DeFragS: Mitigating Resource Fragmentation in GPU Clusters Through Spatial-Temporal Scheduling.
.- OA-WGAN: A Clustering-Guided GAN for Disk Fault Augmentation.
.- Dynamic Adaptive Fault-Tolerance in Stream Computing Systems under Resource Constraints.
.- Auto-CLOUDSC: An Auto-Generation Framework for Vectorization and Optimization of Cloud Microphysics Parameterization on ARM CPUs.
.- Coalition Formation-Based Auction for Deep Neural Network Inference in Vehicular Edge Computing.
.- FedCSAD: Federated Learning with Contextual Client Selection and Confidence-Weighted Multi-Teacher Knowledge Distillation in Power Equipment Inspection.
.- DRL-Based Collaborative Caching Strategy Considering User Preference Prediction in UAV-Assisted Vehicular Edge Computing.
.- MPPTS: Multi-Factor Predictive Priority Task Scheduling Algorithm for Heterogeneous Systems.
.- FPAMM: Fine-Grained Pipeline Architecture Accelerator for the Novel Transformer Architecture - Monarch Mixer.
.- Ly-MAPPO: Enhancing Dynamic V2V Communication via Lyapunov-Based MAPPO under Multi-Dimensional Constraints.
.- A Multi-Strategy Communication Optimization and Adaptive Model Splitting Scheme for Federated Split Learning.
.- Multi-Modal Parallelism Scheduling for Heterogeneous Multicore Computing Systems.
.- A Lightweight Semantic RGB-D vSLAM for Environments with Dynamic Rigid Objects.
.- RAFL-DSV: Robustness-Driven Adaptive Federated Learning with Dynamic Shapley Value.
.- MVDR: Enabling Transaction Dependency Repair in Interactive Transaction Processing.
.- Vectorized Optimization Implementation of Multi-Scalar Multiplication Based on Heterogeneous Digital Signal Processor.
.- FedHyperClass: Boosting Cross-Modal Consensus for Federated Learning with Unimodal Clients.
.- DisDiffAD: A Distributed Diffusion-Based Framework for Efficient Time Series Anomaly Detection in Edge-Cloud Environment.