Search Schedule »

Show All »

Timezone: America/Los_Angeles »

Indicates the beginning of a session

 Sun Aug 28, 2022

Time Mission City Lobby
04:00 PM (Registration Desks)
Registration Check-in Desk

 Mon Aug 29, 2022

Time Room 203 Room 204 Exhibit Hall A Mission City Ballroom Mission City Lobby
07:00 AM (Registration Desks)
Registration Check-in Desk
08:45 AM (Orals)
Outstanding Papers
Pathways: Asynchronous Distributed Dataflow for ML
09:03 AM (Orals)
QuadraLib: A Performant Quadratic Neural Network Library for Architecture Optimization and Design Exploration
09:21 AM (Orals)
Random Offset Block Embedding (ROBE) for compressed embedding tables in deep learning recommendation systems
09:39 AM (Orals)
ML-EXray: Visibility into ML Deployment on the Edge
09:57 AM (Orals)
GPU Semiring Primitives for Sparse Neighborhood Methods
10:30 AM (Invited Talks)
Systems for ML and ML for Systems: A Virtuous Cycle
11:30 AM (Panels)
The Future of MLSys: Industry and Academia - panel
01:00 PM (Orals, Tutorials)
Time Series Anomaly Detection : Tools, Techniques & Tricks
Gemmini: Enabling Systematic Deep-Learning Architecture Evaluation via Full-Stack Integration
Distributed and Parallel Learning
BNS-GCN: Efficient Full-Graph Training of Graph Convolutional Networks with Partition-Parallelism and Random Boundary Node Sampling
01:18 PM (Orals)
Sequential Aggregation and Rematerialization: Distributed Full-batch Training of Graph Neural Networks on Large Graphs
01:36 PM (Orals)
PAPAYA: Practical, Private, and Scalable Federated Learning
01:54 PM (Orals)
LightSecAgg: a Lightweight and Versatile Design for Secure Aggregation in Federated Learning
02:15 PM (Orals)
ML Compilers & Runtime
The CoRa Tensor Compiler: Compilation for Ragged Tensors with Minimal Padding
02:33 PM (Orals)
Apollo: Automatic Partition-based Operator Fusion through Layer by Layer Optimization
02:51 PM (Orals)
DietCode: Automatic Optimization for Dynamic Tensor Programs
03:09 PM (Orals)
Bit-serial Weight Pools: Compression and Arbitrary Precision Execution of Neural Networks on Resource Constrained Processors
04:00 PM (Orals)
Systems for ML 1
MLPerf Mobile Inference Benchmark: An Industry-Standard Open-Source Machine Learning Benchmark for On-Device AI
04:18 PM (Orals)
TorchSparse: Efficient Point Cloud Inference Engine
04:36 PM (Orals)
Hydrozoa: Dynamic Hybrid-Parallel DNN Training on Serverless Containers
04:54 PM (Orals)
Plumber: Diagnosing and Removing Performance Bottlenecks in Machine Learning Data Pipelines
05:12 PM (Orals)
Graphiler: Optimizing Graph Neural Networks with Message Passing Data Flow Graph
05:30 PM (Receptions)
Welcome Reception

 Tue Aug 30, 2022

Time Room 204 Exhibit Hall A Mission City Lobby
06:30 AM (Registration Desks)
Registration Check-in Desk
08:00 AM (Tutorials)
ML-based Computer System Telemetry Analytics
08:45 AM (Orals)
Systems for ML 2
QuClassi: A Hybrid Deep Neural Network Architecture based on Quantum State Fidelity
09:03 AM (Orals)
VirtualFlow: Decoupling Deep Learning Models from the Underlying Hardware
09:21 AM (Orals)
TAGLETS: A System for Automatic Semi-Supervised Learning with Auxiliary Data
09:39 AM (Orals)
mmSampler: Efficient Frame Sampler for Multimodal Video Retrieval
09:57 AM (Orals)
Sustainable AI: Environmental Implications, Challenges and Opportunities
10:30 AM (Invited Talks)
Towards Building a Responsible Data Economy
01:00 PM (Orals, Tutorials)
Sparsity in ML: Understanding and Optimizing Sparsity in Neural Networks Running on Heterogeneous Systems
ML for Systems
Matchmaker: Data Drift Mitigation in Machine Learning for Large-Scale Systems
01:18 PM (Orals)
A Transferable Approach for Partitioning Machine Learning Models on Multi-Chip-Modules
01:36 PM (Orals)
SLA-Driven ML INFERENCE FRAMEWORK FOR CLOUDS WITH HETEROGENEOUS ACCELERATORS
01:54 PM (Orals)
NURD: Negative-Unlabeled Learning for Online Datacenter Straggler Prediction
02:15 PM (Orals)
Hardware Efficient ML
Collapsible Linear Blocks for Super-Efficient Super Resolution
02:33 PM (Orals)
Towards the Co-design of Neural Networks and Accelerators
02:51 PM (Orals)
On the Utility of Gradient Compression in Distributed Training Systems
03:09 PM (Orals)
Efficient Strong Scaling Through Burst Parallel Training
04:00 PM (Orals)
Testing, Debugging and Monitoring & Security
Revelio: ML-Generated Debugging Queries for Finding Root Causes in Distributed Systems
04:18 PM (Orals)
Randomness in Neural Network Training: Characterizing the Impact of Tooling
04:36 PM (Orals)
Synthesizing Optimal Parallelism Placement and Reduction Strategies on Hierarchical Systems for Deep Learning
04:54 PM (Orals)
dPRO: A Generic Performance Diagnosis and Optimization Toolkit for Expediting Distributed DNN Training
05:12 PM (Orals)
A Tale of Two Models: Constructing Evasive Attacks on Edge Models

 Wed Aug 31, 2022

Time Room 203 Room 204 Exhibit Hall A Mission City Lobby
06:30 AM (Registration Desks)
Registration Check-in Desk
08:00 AM (Tutorials)
Online Experimentation for Cloud Applications
08:45 AM (Orals)
Efficient Training
REX: Revisiting Budgeted Training with an Improved Schedule
09:03 AM (Orals)
SRIFTY: Swift and Thrifty Distributed Neural Network Training on the Cloud
09:21 AM (Orals)
Accelerating Training and Inference of Graph Neural Networks with Fast Sampling and Pipelining
09:39 AM (Orals)
Improving Model Training with Multi-fidelity Hyperparameter Evaluation
09:57 AM (Orals)
Gyro Dropout: Maximizing Ensemble Effect in Neural Network Training
10:30 AM (Invited Talks)
Accelerating Engineering with Machine Learning
01:00 PM (Orals, Tutorials)
ASTRA-sim: Enabling SW/HW Co-Design Exploration for Distributed Deep Learning Training Platforms
Training-Free Approaches for Edge AI: Challenges, Opportunities and Progress
ML Programming Models and Abstractions & Interpretability and Explainability of ML
Understanding GNN Computational Graph: A Coordinated Computation, IO, and Memory Perspective
01:18 PM (Orals)
torch.fx: Practical Program Capture and Transformation for Deep Learning in Python
01:36 PM (Orals)
FROTE: Feedback Rule-Driven Oversampling for Editing Models
01:54 PM (Orals)
TyXe: Pyro-based Bayesian neural nets for Pytorch
02:15 PM (Orals)
Efficient Inference and Model Serving
ULPPACK: Fast Sub-8-bit Matrix Multiply on Commodity SIMD Hardware
02:33 PM (Orals)
AccMPEG: Optimizing Video Encoding for Accurate Video Analytics
02:51 PM (Orals)
HALOS: Hashing Large Output Space for Cheap Inference
03:09 PM (Orals)
Learning Compressed Embeddings for On-Device Inference
03:27 PM (Orals)
Bolt: Bridging the Gap between Auto-tuners and Hardware-native Performance
03:45 PM (Orals)
URSABench: A System for Comprehensive Benchmarking of Bayesian Deep Neural Network Models and Inference methods

 Thu Sep 01, 2022

Time Mission Ballroom MR2 Mission Ballroom MR1 Mission Ballroom MR3 Mission Ballroom B4 Mission Ballroom B5 Mission City Lobby
07:00 AM (Registration Desks)
Registration Check-in Desk
08:00 AM (Symposiums, Workshops)
Cross-Community Federated Learning: Algorithms, Systems and Co-designs
Practical Adoption Challenges of ML for Systems in Industry (PACMI)
Chips & Compilers
Benchmarking Machine Learning Workloads on Emerging Hardware
Cloud Intelligence: AI/ML for Efficient and Manageable Cloud Services