USENIX ATC '24 and OSDI '24

Attending this event?

USENIX ATC '24 and OSDI '24

7:30am PDT

Badge Pickup Lobby West

8:00am PDT

Continental Breakfast Grand Ballroom Foyer

9:00am PDT

Joint Keynote Address: Scaling AI Sustainably: An Uncharted Territory Grand Ballroom ABGH

10:00am PDT

Break with Refreshments Grand Ballroom Foyer

10:30am PDT

OSDI ’24 Opening Remarks and Awards Grand Ballroom ABGH USENIX ATC ’24 Opening Remarks, Awards, and Presentation of the 2024 USENIX Lifetime Achievement (Flame) Award Grand Ballroom CD

10:45am PDT

Sabre: Hardware-Accelerated Snapshot Compression for Serverless MicroVMs Grand Ballroom ABGH Harmonizing Efficiency and Practicability: Optimizing Resource Utilization in Serverless Computing with Jiagu Grand Ballroom CD Power-aware Deep Learning Model Serving with μ-Serve Grand Ballroom EF

11:05am PDT

Nomad: Non-Exclusive Memory Tiering via Transactional Page Migration Grand Ballroom ABGH

11:10am PDT

ALPS: An Adaptive Learning, Priority OS Scheduler for Serverless Functions Grand Ballroom CD Fast Inference for Probabilistic Graphical Models Grand Ballroom EF

11:25am PDT

Managing Memory Tiers with CXL in Virtualized Environments Grand Ballroom ABGH

11:35am PDT

Starburst: A Cost-aware Scheduler for Hybrid Cloud Grand Ballroom CD Cost-Efficient Large Language Model Serving for Multi-turn Conversations with CachedAttention Grand Ballroom EF

11:45am PDT

Harvesting Memory-bound CPU Stall Cycles in Software with MSH Grand Ballroom ABGH

12:00pm PDT

StreamBox: A Lightweight GPU SandBox for Serverless Inference Workflow Grand Ballroom CD PUZZLE: Efficiently Aligning Large Language Models through Light-Weight Context Switch Grand Ballroom EF

12:05pm PDT

A Tale of Two Paths: Toward a Hybrid Data Plane for Efficient Far-Memory Applications Grand Ballroom ABGH

12:25pm PDT

DRust: Language-Guided Distributed Shared Memory with Fine Granularity, Full Transparency, and Ultra Efficiency Grand Ballroom ABGH ATC Conference Luncheon Santa Clara Ballroom

12:45pm PDT

OSDI Conference Luncheon Santa Clara Ballroom

2:00pm PDT

Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve Grand Ballroom ABGH ScalaAFA: Constructing User-Space All-Flash Array Engine with Holistic Designs Grand Ballroom CD PeRF: Preemption-enabled RDMA Framework Grand Ballroom EF

2:20pm PDT

ServerlessLLM: Low-Latency Serverless Inference for Large Language Models Grand Ballroom ABGH

2:25pm PDT

FastCommit: resource-efficient, performant and cost-effective file system journaling Grand Ballroom CD CyberStar: Simple, Elastic and Cost-Effective Network Functions Management in Cloud Network at Scale Grand Ballroom EF

2:40pm PDT

InfiniGen: Efficient Generative Inference of Large Language Models with Dynamic KV Cache Management Grand Ballroom ABGH

2:50pm PDT

ZMS: Zone Abstraction for Mobile Flash Storage Grand Ballroom CD OSMOSIS: Enabling Multi-Tenancy in Datacenter SmartNICs Grand Ballroom EF

3:00pm PDT

Llumnix: Dynamic Scheduling for Large Language Model Serving Grand Ballroom ABGH

3:15pm PDT

Ethane: An Asymmetric File System for Disaggregated Persistent Memory Grand Ballroom CD ETC: An Elastic Transmission Control Using End-to-End Available Bandwidth Perception Grand Ballroom EF

3:20pm PDT

DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving Grand Ballroom ABGH

3:40pm PDT

Break with Refreshments Grand Ballroom Foyer

4:10pm PDT

ACCL+: an FPGA-Based Collective Engine for Distributed Applications Grand Ballroom ABGH More is Different: Prototyping and Analyzing a New Form of Edge Server with Massive Mobile SoCs Grand Ballroom CD Limitations and Opportunities of Modern Hardware Isolation Mechanisms Grand Ballroom EF

4:30pm PDT

Beaver: Practical Partial Snapshots for Distributed Cloud Services Grand Ballroom ABGH

4:35pm PDT

HiP4-UPF: Towards High-Performance Comprehensive 5G User Plane Function on P4 Programmable Switches Grand Ballroom CD FetchBPF: Customizable Prefetching Policies in Linux with eBPF Grand Ballroom EF

4:50pm PDT

Fast and Scalable In-network Lock Management Using Lock Fission Grand Ballroom ABGH

5:00pm PDT

KEPC-Push: A Knowledge-Enhanced Proactive Content Push Strategy for Edge-Assisted Video Feed Streaming Grand Ballroom CD Fast (Trapless) Kernel Probes Everywhere Grand Ballroom EF

5:10pm PDT

Chop Chop: Byzantine Atomic Broadcast to the Network Limit Grand Ballroom ABGH

5:25pm PDT

High-density Mobile Cloud Gaming on Edge SoC Clusters Grand Ballroom CD HydraRPC: RPC in the CXL Era Grand Ballroom EF

5:40pm PDT

ExtMem: Enabling Application-Aware Virtual Memory Management for Data-Intensive Applications Grand Ballroom EF

6:00pm PDT

OSDI ’24 Poster Session and Reception Santa Clara Ballroom

7:30pm PDT

Joint Student Meet-up Bayshore Room Google Sponsor Event: Ask-Me-Anything Panel Session Magnolia Room Birds-of-a-Feather Sessions (BoFs) Central Room, Tasman Room

8:00am PDT

Continental Breakfast Grand Ballroom Foyer Badge Pickup Lobby West

9:00am PDT

Enabling Tensor Language Model to Assist in Generating High-Performance Tensor Programs for Deep Learning Grand Ballroom ABGH Telescope: Telemetry for Gargantuan Memory Footprint Applications Grand Ballroom CD WingFuzz: Implementing Continuous Fuzzing for DBMSs Grand Ballroom EF

9:20am PDT

Ladder: Enabling Efficient Low-Precision Deep Learning Computing through Hardware-aware Tensor Transformation Grand Ballroom ABGH

9:25am PDT

An Empirical Study of Rust-for-Linux: The Success, Dissatisfaction, and Compromise Grand Ballroom CD Balancing Analysis Time and Bug Detection: Daily Development-friendly Bug Detection in Linux Grand Ballroom EF

9:40am PDT

Caravan: Practical Online Learning of In-Network ML Models with Labeling Agents Grand Ballroom ABGH

9:50am PDT

Scalable and Effective Page-table and TLB management on NUMA Systems Grand Ballroom CD Kivi: Verification for Cluster Management Grand Ballroom EF

10:00am PDT

nnScaler: Constraint-Guided Parallelization Plan Generation for Deep Learning Training Grand Ballroom ABGH

10:15am PDT

UniMem: Redesigning Disaggregated Memory within A Unified Local-Remote Memory Hierarchy Grand Ballroom CD Monarch: A Fuzzing Framework for Distributed File Systems Grand Ballroom EF

10:20am PDT

ChameleonAPI: Automatic and Efficient Customization of Neural Networks for ML Applications Grand Ballroom ABGH

10:40am PDT

Break with Refreshments Grand Ballroom Foyer

11:10am PDT

SquirrelFS: using the Rust compiler to check file-system crash consistency Grand Ballroom ABGH Accelerating the Training of Large Language Models using Efficient Activation Rematerialization and Optimal Hybrid Parallelism Grand Ballroom CD A Secure, Fast, and Resource-Efficient Serverless Platform with Function REWIND Grand Ballroom EF

11:30am PDT

High-throughput and Flexible Host Networking for Accelerated Computing Grand Ballroom ABGH

11:35am PDT

Metis: Fast Automatic Distributed Training on Heterogeneous GPUs Grand Ballroom CD SimEnc: A High-Performance Similarity-Preserving Encryption Approach for Deduplication of Encrypted Docker Images Grand Ballroom EF

11:50am PDT

IntOS: Persistent Embedded Operating System and Language Support for Multi-threaded Intermittent Computing Grand Ballroom ABGH

12:00pm PDT

FwdLLM: Efficient Federated Finetuning of Large Language Models with Perturbed Inferences Grand Ballroom CD mmTLS: Scaling the Performance of Encrypted Network Traffic Inspection Grand Ballroom EF

12:10pm PDT

Data-flow Availability: Achieving Timing Assurance in Autonomous Systems Grand Ballroom ABGH

12:25pm PDT

ATC Conference Luncheon Santa Clara Ballroom

12:30pm PDT

Microkernel Goes General: Performance and Compatibility in the HongMeng Production Microkernel Grand Ballroom ABGH

12:50pm PDT

OSDI Conference Luncheon Santa Clara Ballroom

2:00pm PDT

When will my ML Job finish? Toward providing Completion Time Estimates through Predictability-Centric Scheduling Grand Ballroom ABGH Pecan: Cost-Efficient ML Data Preprocessing with Automatic Transformation Ordering and Hybrid Placement Grand Ballroom CD QDSR: Accelerating Layer-7 Load Balancing by Direct Server Return with QUIC Grand Ballroom EF

2:20pm PDT

Optimizing Resource Allocation in Hyperscale Datacenters: Scalability, Usability, and Experiences Grand Ballroom ABGH

2:25pm PDT

OPER: Optimality-Guided Embedding Table Parallelization for Large-scale Recommendation Model Grand Ballroom CD Evaluating Chiplet-based Large-Scale Interconnection Networks via Cycle-Accurate Packet-Parallel Simulation Grand Ballroom EF

2:40pm PDT

μSlope: High Compression and Fast Search on Semi-Structured Logs Grand Ballroom ABGH

2:50pm PDT

MAGPY: Compiling Eager Mode DNN Programs by Monitoring Execution States Grand Ballroom CD Config-Snob: Tuning for the Best Configurations of Networking Protocol Stack Grand Ballroom EF

3:00pm PDT

ServiceLab: Preventing Tiny Performance Regressions at Hyperscale through Pre-Production Testing Grand Ballroom ABGH

3:15pm PDT

Quant-LLM: Accelerating the Serving of Large Language Models via FP6-Centric Algorithm-System Co-Design on Modern GPUs Grand Ballroom CD Conspirator: SmartNIC-Aided Control Plane for Distributed ML Workloads Grand Ballroom EF

3:20pm PDT

MAST: Global Scheduling of ML Training across Geo-Distributed Datacenters at Hyperscale Grand Ballroom ABGH

3:40pm PDT

Break with Refreshments Grand Ballroom Foyer

4:10pm PDT

Automatically Reasoning About How Systems Code Uses the CPU Cache Grand Ballroom ABGH FBMM: Making Memory Management Extensible With Filesystems Grand Ballroom CD SuperBench: Improving Cloud AI Infrastructure Reliability with Proactive Validation Grand Ballroom EF

4:30pm PDT

VeriSMo: A Verified Security Module for Confidential VMs Grand Ballroom ABGH

4:35pm PDT

Mangosteen: Fast Transparent Durability for Linearizable Applications using NVM Grand Ballroom CD Removing Obstacles before Breaking Through the Memory Wall: A Close Look at HBM Errors in the Field Grand Ballroom EF

4:50pm PDT

Validating the eBPF Verifier via State Embedding Grand Ballroom ABGH

5:00pm PDT

FlexMem: Adaptive Page Profiling and Migration for Tiered Memory Grand Ballroom CD MSFRD: Mutation Similarity based SSD Failure Rating and Diagnosis for Complex and Volatile Production Environments Grand Ballroom EF

5:10pm PDT

Using Dynamically Layered Definite Releases for Verifying the RefFS File System Grand Ballroom ABGH

5:30pm PDT

Anvil: Verifying Liveness of Cluster Management Controllers Grand Ballroom ABGH

6:00pm PDT

USENIX ATC ’24 Poster Session and Reception Santa Clara Ballroom

7:30pm PDT

Databricks Sponsor Event Bayshore Room Birds-of-a-Feather Sessions (BoFs) Central Room, Tasman Room

8:30pm PDT

Futurewei Sponsor Event Bayshore Room

8:00am PDT

Continental Breakfast Grand Ballroom Foyer Badge Pickup Lobby West

9:00am PDT

DSig: Breaking the Barrier of Signatures in Data Centers Grand Ballroom ABGH Diagnosing Application-network Anomalies for Millions of IPs in Production Clouds Grand Ballroom CD Panorama: Optimizing Internet-scale Users’ Routes from End to End Grand Ballroom EF

9:20am PDT

Ransom Access Memories: Achieving Practical Ransomware Protection in Cloud with DeftPunk Grand Ballroom ABGH

9:25am PDT

Data Caching for Enterprise-Grade Petabyte-Scale OLAP Grand Ballroom CD Enhancing Resource Management of the World's Largest PCDN System for On-Demand Video Streaming Grand Ballroom EF

9:40am PDT

Secret Key Recovery in a Global-Scale End-to-End Encryption System Grand Ballroom ABGH

9:50am PDT

Full Lifecycle Data Analysis on a Large-scale and Leadership Supercomputer: What Can We Learn from It? Grand Ballroom CD TileClipper: Lightweight Selection of Regions of Interest from Videos for Traffic Surveillance Grand Ballroom EF

10:00am PDT

Flock: A Framework for Deploying On-Demand Distributed Trust Grand Ballroom ABGH

10:15am PDT

ATC Break with Refreshments Grand Ballroom Foyer

10:20am PDT

OSDI Break with Refreshments Grand Ballroom Foyer

10:50am PDT

FairyWREN: A Sustainable Cache for Emerging Write-Read-Erase Flash Interfaces Grand Ballroom ABGH Expeditious High-Concurrency MicroVM SnapStart in Persistent Memory with an Augmented Hypervisor Grand Ballroom CD Efficient Decentralized Federated Singular Vector Decomposition Grand Ballroom EF

11:10am PDT

Massively Parallel Multi-Versioned Transaction Processing Grand Ballroom ABGH

11:15am PDT

Taming Hot Bloat Under Virtualization with HUGESCOPE Grand Ballroom CD Models on the Move: Towards Feasible Embedded AI for Intrusion Detection on Vehicular CAN Bus Grand Ballroom EF

11:30am PDT

Burstable Cloud Block Storage with Data Processing Units Grand Ballroom ABGH

11:40am PDT

CrossMapping: Harmonizing Memory Consistency in Cross-ISA Binary Translation Grand Ballroom CD CPC: Flexible, Secure, and Efficient CVM Maintenance with Confidential Procedure Calls Grand Ballroom EF

11:50am PDT

Motor: Enabling Multi-Versioning for Distributed Transactions on Disaggregated Memory Grand Ballroom ABGH

12:05pm PDT

ATC Lunch (on your own) N/A

12:10pm PDT

OSDI Lunch (on your own) N/A

1:40pm PDT

Detecting Logic Bugs in Database Engines via Equivalent Expression Transformation Grand Ballroom ABGH RL-Watchdog: A Fast and Predictable SSD Liveness Watchdog on Storage Systems Grand Ballroom CD gVulkan: Scalable GPU Pooling for Pixel-Grained Rendering in Ray Tracing Grand Ballroom EF

2:00pm PDT

Inductive Invariants That Spark Joy: Using Invariant Taxonomies to Streamline Distributed Protocol Proofs Grand Ballroom ABGH

2:05pm PDT

Exploit both SMART Attributes and NAND Flash Wear Characteristics to Effectively Forecast SSD-based Storage Failures in Clusters Grand Ballroom CD vFPIO: A Virtual I/O Abstraction for FPGA-accelerated I/O Devices Grand Ballroom EF

2:20pm PDT

Performance Interfaces for Hardware Accelerators Grand Ballroom ABGH

2:30pm PDT

StreamCache: Revisiting Page Cache for File Scanning on Fast Storage Devices Grand Ballroom CD ScalaCache: Scalable User-Space Page Cache Management with Software-Hardware Coordination Grand Ballroom EF

2:40pm PDT

IronSpec: Increasing the Reliability of Formal Specifications Grand Ballroom ABGH

2:55pm PDT

Scalable Billion-point Approximate Nearest Neighbor Search Using SmartSSDs Grand Ballroom CD Centimani: Enabling Fast AI Accelerator Selection for DNN Training with a Novel Performance Predictor Grand Ballroom EF

3:00pm PDT

Identifying On-/Off-CPU Bottlenecks Together with Blocked Samples Grand Ballroom ABGH

3:20pm PDT

Break with Refreshments Grand Ballroom Foyer

3:40pm PDT

dLoRA: Dynamically Orchestrating Requests and Adapters for LoRA LLM Serving Grand Ballroom ABGH A Difference World: High-performance, NVM-invariant, Software-only Intermittent Computation Grand Ballroom CD

4:00pm PDT

Parrot: Efficient Serving of LLM-based Applications with Semantic Variable Grand Ballroom ABGH

4:05pm PDT

Efficient Large Graph Processing with Chunk-Based Graph Representation Model Grand Ballroom CD

4:20pm PDT

USHER: Holistic Interference Avoidance for Resource Optimized ML Inference Grand Ballroom ABGH

4:30pm PDT

SlimArchive: A Lightweight Architecture for Ethereum Archive Nodes Grand Ballroom CD

4:40pm PDT

Fairness in Serving Large Language Models Grand Ballroom ABGH

4:55pm PDT

Every Mapping Counts in Large Amounts: Folio Accounting Grand Ballroom CD

5:00pm PDT

MonoNN: Enabling a New Monolithic Optimization Space for Neural Network Inference Tasks on Modern GPU-Centric Architectures Grand Ballroom ABGH

5:10pm PDT

USENIX ATC ’24 Closing Remarks Grand Ballroom CD

5:20pm PDT

OSDI ’24 Closing Remarks Grand Ballroom ABGH