Accepted Papers

• Generating representative macrobenchmark microservice systems from distributed traces with ShortKut
• Between Promise and Pain: The Reality of Automating Microservices Diagnosis with Large Language Models
• Towards Fully Disaggregated Recommendation Model Serving
• Guess Who’s Back: Comeback Interval based Multi-tier Memory Management
• Autokernel: ML-native Dataplane Operating Systems
• TARDIS: A GPU-Centric KV Cache Service for Efficient LLM Inference
• Indispensable CPU-centric Checkpointing for GPUs
• Cloud abstractions for AI workloads
• Towards Microsecond-Scale VM Core Provisioning Agility on Serverless Platforms
• Meshilon: A Merging-based Java Garbage Collector
• CrazyOS: Memory management and scheduling should operate through a unified subsystem to improve maintainability
• Staying in the Zone - Retrofitting Zoned Storage into a Scalable Enterprise File System
• SHOC: A Simplified Programming Model for Hardware Offloading on DPUs
• Implementing a persistent key-value store in a tamper-resistant device for SGX enclave applications
• FPGAs are the Hero In-Network Computing Needs
• Mainframe-style channel controllers for modern disaggregated memory systems
• Metadata-Driven Near-Exclusive Caching for NVMe-oF SANs with eBPF
• Can LLMs Replace Time-Tested System Policies? Perhaps
• A System-level Abstraction and Service for Flourishing AI-powered Applications
• CPU Autoscaling With a Kernel of Truth
• Nesting Overlay File Systems with ShadowWhiteout
• HyperGen: Optimizing Generative Inference with Long Prompts for Resource-Constrained Systems
• Parsec: Fast, Scalable, and Secure Design with Wait-Free Parallelism
• Chimera-VDB: Mixed-Precision Vector Database with HNSW Index for RAG-LLM
• Exploring B+-Tree Implementations Using Scratchpad Memory
• Using Recursive Attestation to Scale Trust in Modern Heterogeneous Cloud Architectures