Accepted papers - main conference
- — Exploring Memory Partitioning of LLMs for Faster Inference on CPUs
- — Emulation vs Simulation: A Case Study from Developing Congestion Control Algorithms in Low Earth Orbit Satellite Networks
- — Decentralized Edge Caching under Budget and Storage Constraints: A Game-Theoretic Approach
- — Evaluating MFU as a Portable Proxy for GPU Power in Energy-Aware LLM Simulation
- — Understanding Large Payload Overhead in SQLite: Analysis and Design Implications
- — InFactPlanner: Planning Sustainable Geo-Distributed LLM Data Centers
- — CSDPaxos: Computational Storage Device-Accelerated Paxos Systems with SmartSSD
- — OMNI: OCS Middleware for Network Improvement
- — Enabling Zero-Knowledge Proving of Large Language Models on a Consumer GPU
- — Time-to-Consistency in Saga Transactions: An Analytical Model and Empirical Validation
- — Complex Quantum G-Networks with Product Form
- — quPIM: Enabling and Characterizing Quantum Circuit Simulation on Commercial Processing-in-Memory Systems
- — Accurate Simulation of Distributed Training Jobs with Network Contention Modeling
- — MCP-GRANITE Benchmark: GRANularity Interface TEsting for MCP-Based LLM Agents
- — Island-Ready Communication Networks: Connectivity Monitoring and State Synchronization for 5G Crisis Connectivity at the Edge
- — Three-Dimensional Trajectory and Task-Offloading Optimization in Blockchain-Assisted Multi-UAV Edge Computing
- — Adaptive Direction-Aware Fault-Tolerant Routing in Data Center Networks
- — DeepReasonRCA: Real-Time Root Cause Analysis via Projection-Based Reconstruction Reasoning
- — Input Resolution Matters: Real-Time Object Detection Latency
- — Measuring Compression Honestly: A Dense-FP16 Baseline for Quantization and Sparsity on CNNs and Vision Transformers
- — HaSP: Harnessing Storage-side Parallelism for Matrix Multiplication
- — Xronos: Heterogeneity-Aware Tensor Parallelism for Collaborative LLM Fine-Tuning on Edge CPUs
- — Multi-Agent Hierarchical Reinforcement Learning for Runtime Optimization of AI Pipelines in Smart Eyewear
- — Not All AI Agents Are Equal: Characterizing Resource and Performance Dynamics
- — PAVE: Partial KV Cache Eviction for Fine-Grained KV Reclamation in LLM Serving
Accepted papers - MASCOTS workshop
- — Comparative Characterization of KV Cache Management Strategies for LLM Inference
- — Decision-Guided Clustering in the Probability Space for Energy Efficiency in LoRaWAN Networks
- — Requested Wall Time Across Public HPC Workload Traces: Drift and Limits of Reuse
- — MetaGC: Improving SSD Garbage Collection with Randomization
- and Piotr Czekalski — A Data-driven Queueing-Theoretic Performance Modelling Framework for Plugged-in Electric Vehicle (PEV) Fast-Charging Stations
- — On Truncated Attachment Mechanism in Social Networks
- — Aladdin: Joint Placement and Scaling for SLO-Aware LLM Serving
- — A Flexible and High-Performance Simulator for Multiserver Job Queuing Model
- — Evaluating the Transfer of RAG Poisoning Defenses to Multi-Hop Question Answering
- — Split-Step Channel Models for Timed Security Protocols: Verification Precision and Cost at Scale
- — Heuristic Decentralized Orchestration of Energy- and Time-constrained Heterogeneous Satellite Computing Services
- — Investigating the Impact of Inter-Arrival Time and Service Time on the Performance of H2/H2/k
- — G-Network Based Optimisation of Coupled Pedestrian and Information Flows in High-Occupancy Public Spaces
Technically Co-Sponsored by the IEEE Computer Society
