Detailed Program

See the Schedule Overview for the full week and rooms.
08:00
Registration open
09:00 – 10:30
TSCSRoom: CC I
Trustworthy and Secure Cloud Services
Full program on the workshop website.
10:30 – 11:00
Coffee Break
11:00 – 12:30
12:30 – 14:00
Lunch (Hotel Restaurant)
14:00 – 16:00
16:00 – 16:30
Coffee Break
16:30 – 17:30
18:30 – 20:30
Welcome Reception (Hotel restaurant)
08:00
Registration open
09:00 – 10:30
Conference Opening and Keynote – Kyle ChardRoom: CC I
10:30 – 11:00
Coffee Break
11:00 – 12:30
UCC-1Room: CC I
Novel runtime models for distributed AI
11:00–11:20Inference Pipeline Serving with In-Place Vertical Scaling on Spatially Shared CPUs and GPUsMehran Salmani, Kamran Razavi, Peter Amthor and Boris Koldehofe
11:20–11:40JABS: Jointly Adapting Batch Size and GPU SM Allocation for Deadline-Aware ML InferenceTheo Radig and Holger Karl
11:40–12:00Enabling Full Support for Heterogeneous Platforms in StarPU’s Server-Client Execution ModelNicolas Vanz, Vanderlei Munhoz Pereira Filho, Márcio Castro, Laércio L. Pilla and Olivier Aumage
12:00–12:20Porting a High-Level Synthesis Dynamic Programming Framework to HBM-Enabled Cloud FPGAs: An Early Study on AWS EC2 F2Roberto Chura Chambi, Cristina Boeres and Vinod Rebello
12:20–12:30ScaleStream: Adaptive Locality-Aware Key Bucketing for Stateful Stream ProcessingPritish Mishra, Alexandre da Silva Veith, Oana Balmau and Eyal de Lara
12:30 – 14:00
Lunch (Hotel Restaurant)
14:00 – 16:00
UCC-2Room: CC I
Distributed and Compound AI
14:00–14:20Clockwork: Scheduling Compound AI Systems in Space Data CentersMilos Gravara, Andrija Stanisic and Stefan Nastic
14:20–14:40EdgeLLMServe: Distributing LLM models on Heterogeneous Edge/Fog Cluster EnvironmentTruong-Thanh Le, Hoang-Loc La, Amir Taherkordi, Frank Eliassen, Phuong Hoai Ha and Peiyuan Guan
14:40–15:00Pine: Hierarchical Projection Pruning for Resource-Efficient Mixture-of-Experts LLMsAbdullah Nazari, Bailey Eccles and Blesson Varghese
15:00–15:20Reshaping Sparse Neural Networks for Communication-Efficient Distributed Edge InferenceWoohyeok Jang and Seung-Jae Han
15:20–15:40Communication Efficiency in Quantum Federated Learning Depends on Optimizer Geometry and System ScaleRómulo Walter Condori Bustincio, Allan M. de Souza and Luiz F. Bittencourt
16:00 – 16:30
Coffee Break
16:30 – 18:00
UCC-3Room: CC I
Networking and Network Management
16:30–16:50Packet-Loss-Aware Elastic Slicing for 5G Open RAN: User Satisfaction EvaluationCarlos Lima, William Nascimento, Ricardo Silva, Maykon Silva, Francisco Hugo Neto, Dario Vieira, Augusto Neto, Antonio Campos and Vicente Sousa
16:50–17:10PATH: Path-aware Adaptive Traffic at the Host for Load Balancing High Performance Data TransfersDaniel José Ventorim Nunes, Magnos Martinello, Cristina Klippel Dominicini, Everson Scherrer Borges and Rafael Silva Guimarães
17:10–17:30Cloud-Edge Continuum for Telco-Cloud Infrastructures: An Integrated Support System for End-to-End SlicingMathews Phillipp Santos de Lima, Nelson Ion de Oliveira, Rafael Araújo de Andrade, Paulo Eduardo da Silva Jr., Daniel Rodrigues de Luna, Maykon Renan Pereira da Silva, Francisco Hugo Costa Neto, Eduardo Coelho Cerqueira, Denis Lima Rosário, Carlos Alberto Astudillo, Dario Vieira, Augusto José Venâncio Neto and Vicente A. de Sousa Jr.
17:30–17:50Latency-Aware Inference Offloading in a Mobile Cloud-Fog-Edge Continuum for Multi-UAV Maritime Search and RescueWillian Mattos, Eduardo Inacio and Patricia Plentz
17:50–18:00XFlowLyzer: Extreme-Rate eBPF/XDP Feature Extractor for DDoS TelemetryLeonardo Herkenhoff, Ana Martins, Gilmar Vassoler, Rodolfo Villaça, Eduardo Zambon, Matheus Castanho and Cristina Dominicini
08:30
Registration open
09:00 – 10:00
10:00 – 10:30
Coffee Break + Poster Session
10:30 – 12:30
UCC-4Room: CC I
Resource Management
10:30–10:50QoE-Aware Resource Allocation for Edge Computing–Based Cloud GamingManoella Azevedo and Tiago Ferreto
10:50–11:10One Policy, Variable Task Counts: PPO-Based Vertical Scaling for Edge Stream ProcessingGuilherme Santos, Carlos Henrique Kayser and Tiago Ferreto
11:10–11:30Reliability-Aware Adaptive Replica Spreading in Kubernetes SchedulingCaetano Albuquerque, Fabio Morais and Giovanni Farias
11:30–11:50Stochastic selection and memory-aware decomposition for optimization-based workflow placement in serverless computingMoritz Heinlein, Joshua Adamek, Natalie Carl, Trever Schirmer, David Bermbach and Sergio Lucia
11:50–12:10DRUM: Risk- and Uncertainty-Aware Dynamic Deadline Orchestration in the Edge–Cloud ContinuumFrancesco La Rosa, Pierluigi Dell’Acqua, Mario Colosi and Massimo Villari
12:30 – 14:00
Lunch (Hotel Restaurant)
14:00 – 15:00
PanelRoom: CC I
15:00 – 16:00
UCC-5Room: CC I
Security, Trust, Privacy and Policies
15:00–15:20Combining Diversity and Redundancy with Reinstantiation for Moving Target Defense in Fog and Cloud EnvironmentsIago Almeida, Israel Araújo, Luiza Paiva, Juliana Carvalho, Luiz Fernando Bittencourt and Francisco Airton Silva
15:20–15:40Bi-chain: Governance-Bound Cross-Chain Settlement with Bounded Convergence and Cut-Consistent Coordinator-Bias DetectionZeba Mahmood and Stephan Recker
15:40–16:00From Frame to Episode: Evidence-Preserving Alert Mediation for Selective Inference in Vehicular Edge SystemsAngela Albarello, Edna Dias Canedo, Vinícius P. Gonçalves, Leandro A. Villas, Roger Kreutz Immich, Rodolfo I Meneguette and Geraldo P Rocha Filho
16:00 – 16:30
Coffee Break
16:30 – 17:30
Conference Transfer to Gala Dinner
18:30 – 21:00
Conference Gala Dinner
08:30
Registration open
09:00 – 10:30
UCC-6Room: CC I
Deployment Models
09:00–09:20On the Viability of using Simulation to Assess Cloud Offload for HPC WorkloadsOtho José Sirtoli Marcondes, Rayan Raddatz de Matos, Matheus Augusto Tregnago, Pedro Henrique Boniatti Colle and Lucas Schnorr
09:20–09:40Transparent Resilience for Legacy HPC Applications on Cloud Spot Instances with MANAArtur Luiz Rizzato Toru Soda, Vanderlei Munhoz Pereira Filho and Márcio Castro
09:40–10:00Enabling Cost-effective MPI on Cloud Spot Instances via Malleable Execution using RayTingkai Liu, Yue Yuan, Yifei Song, Sara Kokkila-Schumacher, Marquita Ellis and Volodymyr Kindratenko
10:00–10:10A Scalable and Auditable Cloud-Based ETL Pipeline Architecture for Epidemiological Surveillance: A Case Study of Dengue Notifications in Minas Gerais, BrazilKetlen Andrade and Rafael Lima
10:10–10:25SOAR: Rescuing Live Interactive Containers from Spot ReclamationJiasheng Zhuang, Xinghan Chen and Velimir Dedic
UCC-7Room: CC III
Architectural Models and Patterns
09:00–09:20Proactive Contract Unification: A Semantic Similarity Approach for Automated Test Generation in Microservices ArchitecturesManuel Simosa
09:20–09:40AdapBench: Adaptive Early-Stopping and Prioritization for Efficient Microbenchmarking in Cloud-based CI/CDSebastian Koch, Jonas Heisterberg, Nils Japke, Trever Schirmer and David Bermbach
09:40–10:00Understanding Workload Co-location in Virtualized Systems: Performance Isolation and NUMA EffectsArtur Marcon, André Sacilotto, Davi Kniest da Silva, Julio Augusto Borges, Miguel G. Xavier, Tiago C. Ferreto and Cesar A. F. De Rose
10:00–10:10COTTON-NET: Scaling SSI-Based Traceability in the Cotton Supply Chain through Hierarchical Node ArchitectureGilson Sohn Junior, Janaina Fontana Biffi Duarte, Luiz Camargo, Mauricio Pillon, Marcos Simplicio Jr and Charles Miers
10:10–10:25CCL+: Enhancing the Local Cloud-based Collaborative Learning (CCL) Model for Online Data ReductionSaleha Haseeb, Muhammad Usman Joyia, Abdullah Mehmood Minhas, Marcus Liwicki, Fredrik Sandin, Hamam Mokayed and Sergio Martin-del-Campo
10:30 – 11:00
Coffee Break
11:00 – 12:30
UCC-8Room: CC I
Autonomic, Adaptive, Self-management
11:00–11:20Custom Self-Adapter (CSA): A Kubernetes-Native, Application-Aware Self-Adaptation FrameworkHumberto Ribeiro Junior and Nabor Mendonca
11:20–11:40Selection Is Not Adaptation: A Replicated Protocol for Evaluating Action-Constrained DQN Policies in Cloud ConfigurationDammaris Abril Barrera Moreno and Gaddiel Desirena-López
11:40–12:00When a Better Prompt Is Just Noise in LLM-Driven Cloud OrchestrationAndré Cunha, Giovanni Farias and Fabio Morais
12:00–12:10MAPE-SR: A Self-Protecting MAPE-K Based on Security StandardsJorge Soares, Nelson Souto and Fernando Aires
12:10–12:25From Benchmarks to Reality: Why Novelty-Typed Abstention Fails in Metric-Based Cloud RCARaju Chowdhury, Apurv Ghai and Aniruddha Singh Gautam
UCC-9Room: CC III
Energy-Efficiency and Sustainability
11:00–11:20Multi-Profile Power Modeling of Heterogeneous Systems: Insights and LLM Energy PredictionsRohail Gulbaz, Paul Townend and Per-Olov Östberg
11:20–11:40Energy-Efficient Application Placement Techniques for Containerized Edge Computing EnvironmentsRafael Machado Rios and Tiago Coelho Ferreto
11:40–12:00Per-Round Profiling of Energy, Carbon, and Communication in Federated Learning with EnerFLThiago Araújo, Cláudia Brito and Philippe Navaux
12:00–12:20RUNCORE: Energy-Aware Resource Allocation for Hybrid CPU Warehouse SystemsMarcelo Koji Moori, Felipe Nunes Laguna, Hiago Mayk Gomes De Araujo Rocha and Antonio Carlos Schneider Beck
12:30 – 12:45
Closing remarksRoom: CC I
12:45 – 14:00
Lunch (Hotel Restaurant)