- ホーム
- > 洋書
- > ドイツ書
- > Mathematics, Sciences & Technology
- > Computer & Internet
- > hardware
Description
This volume constitutes the proceedings of the 22nd IFIP WG 10.3 International Conference on Network and Parallel Computing, NPC 2026, which was held in Bali, Indonesia, during December 11 13, 2026.
The 70 full papers included in this volume were carefully reviewed and selected from 138 submissions. Topics of interest include, but are not limited to, parallel and distributed applications and algorithms, parallel and distributed architectures and systems, and parallel and distributed software environments and tools.
.- SchedAgent: Useful-Readiness Scheduling for LLM Agent Workflows.
.- RHRS: Fine-Grained Batch Scheduling for Programmable LLM Inference Systems.
.- QB64: Unified Low-Bit Execution for High-Throughput LLM Inference on MT-3000.
.- Efficient Large-Scale Temporal Graph Computing on the GPU.
.- OmniPipe: Adaptive Pipeline Parallelism Scheduling for Efficient Large Language Model Training.
.- TEE-Assisted State-Transition Auditing for Control-Plane Integrity of Black-Box DNN Fingerprint Generation.
.- Dependency-aware Task Scheduling in Collaborative Edge Computing through Reverse Auction.
.- CISelect: Custom Instruction Selection via Multi-Level Cost Modeling Guided by Minimum Description Length.
.- Regime-Aware Dynamic Pricing for Token-Budgeted LMaaS.
.- Parallel Optimization of Sparse Matrix LU Factorization on GPU.
.- Residual Low-Rank Hadamard Reparameterization for Heterogeneous Federated Learning.
.- UNISON: User-Intent and Network-State Orchestrated Framework for 6G Distributed Core Networks.
.- Efficient Wi-Fi CSI-Based Trajectory Tracking on Resource-Constrained Embedded Devices.
.- ASTran: Architecture-Aware Sparse Transformer Inference with Operator Fusion on Ascend NPUs.
.- Bandwidth-Aware Data Placement Routing for Software-Defined Edge Networks.
.- Verifier-Gated LLM Heuristic Evolution for Online Edge-Cluster Scheduling.
.- Accelerating Critical Path Analysis of Large-Scale HPC Traces via Communication Affinity-Aware Scheduling.
.- Beyond Compression Ratio: High-Throughput KV Compression for Disaggregated LLM Serving.
.- CoPrune: Turning Compute Budgets into Deployable Models for TEE-GPU Collaborative Inference.
.- Deploying Terminal LLM Agents on Embedded Edge Devices: Measuring and Mitigating the Agent Scaffold Tax.
.- VineGraph: Parallel and Storage-Efficient Approximate Nearest Neighbor Search via Pre-Navigation Graph.
.- FP4Quant: Efficient 4-Bit Communication Compression for Model-Parallel LLM Training.
.- SemiFE: Resource-Efficient Federated Semi-supervised Learning via Soft Knowledge Aggregation for Heterogeneous Edge Systems.
.- Similarity-Aware Client-Side Reordering for Server-Side Storage Compression in Distributed File Systems.
.- TAISM: A Programming-Model Abstraction for Heterogeneous Trusted Execution Environments.
.- Distributed Multi-Variant Model Placement in Edge Computing: A Game-Theoretic Framework.
.- ST4TE: Selective Synchronization of In-Network Virtual IP Address Caching in Cloud Networks.
.- Parallel Optimization of a Guiding-Center Orbit Simulation Code for Tokamak Plasmas on the Sunway Bluelight II Supercomputer.
.- CA-FEC: Balancing Recovery and Overhead via Context-Aware FEC for Cloud Gaming.
.- Exploiting Profile-Guided Multi-Agent Optimization Workflow for Value-Related Inefficiencies.
.- Azimuth: LLM-Guided Superoptimization with Semantic Search Priors.
.- TetrisServe: Towards Efficient Serving of Agentic LLM Workflows with Demand-Aware Routing.
.- C2-TCCL: Computation-Communication Co-Optimized Path Selection over Heterogeneous Interconnects.
.- PhotonZip: Lifecycle-Aware Compression for HDF5-Based Synchrotron Workflows.
.- An Architecture-Aware Methodology for Optimizing Stencil-Based Applications on the LineShine Supercomputer.
.- RegionZ: Online Reinforcement Learning for Adaptive Multi-page ZRAM Compression in Mobile Devices.
.- MagBound: Relational Legality for Slot-Sanitization Masks in FHE Compilers for Packed CKKS.
.- PEFTS: Multi-Tenant PEFT Serving.
.- NPLB: High-Capacity Stateful Load Balancing on a Novel Programmable Network Processor.
.- ENO-FS: First-Open Native Materialization for On-Demand Container Image Access.
.- FastController: A Control Policy Based on Validated Profiles
