Toggle Poster Visibility
Symposium
Mon Apr 05 08:30 AM -- 05:00 PM (PDT)
Chips and Compilers Symposium
Remarks
Tue Apr 06 08:00 AM -- 08:15 AM (PDT)
Opening Remarks
Invited Talk
Tue Apr 06 08:20 AM -- 09:10 AM (PDT)
Directions for Deep Learning Hardware
Break
Tue Apr 06 09:10 AM -- 09:30 AM (PDT)
Break
Oral
Tue Apr 06 09:30 AM -- 09:50 AM (PDT)
ModularNAS: Towards Modularized and Reusable Neural Architecture Search
[
Paper PDF]
Oral
Tue Apr 06 09:50 AM -- 10:10 AM (PDT)
Fluid: Resource-aware Hyperparameter Tuning Engine
[
Paper PDF]
Oral
Tue Apr 06 10:10 AM -- 10:30 AM (PDT)
MicroNets: Neural Network Architectures for Deploying TinyML Applications on Commodity Microcontrollers
[
Paper PDF]
Oral
Tue Apr 06 10:30 AM -- 10:50 AM (PDT)
Characterizing and Taming Model Instability Across Edge Devices
[
Paper PDF]
Break
Tue Apr 06 10:50 AM -- 11:10 AM (PDT)
Break
Oral
Tue Apr 06 11:10 AM -- 11:30 AM (PDT)
Cortex: A Compiler for Recursive Deep Learning Models
[
Paper PDF]
Oral
Tue Apr 06 11:30 AM -- 11:50 AM (PDT)
A Deep Learning Based Cost Model for Automatic Code Optimization
[
Paper PDF]
Oral
Tue Apr 06 11:50 AM -- 12:10 PM (PDT)
Learning Fitness Functions for Machine Programming
[
Paper PDF]
Oral
Tue Apr 06 12:10 PM -- 12:30 PM (PDT)
CODE: Compiler-based Neuron-aware Ensemble training
[
Paper PDF]
Break
Tue Apr 06 12:30 PM -- 01:30 PM (PDT)
Lunch break
Oral
Tue Apr 06 01:30 PM -- 01:50 PM (PDT)
Pufferfish: Communication-efficient Models At No Extra Cost
[
Paper PDF]
Oral
Tue Apr 06 01:50 PM -- 02:10 PM (PDT)
In-network Aggregation for Shared Machine Learning Clusters
[
Paper PDF]
Oral
Tue Apr 06 02:10 PM -- 02:30 PM (PDT)
Data Movement Is All You Need: A Case Study on Optimizing Transformers
[
Paper PDF]
Oral
Tue Apr 06 02:30 PM -- 02:50 PM (PDT)
Learning on Distributed Traces for Data Center Storage Systems
[
Paper PDF]
Break
Tue Apr 06 02:50 PM -- 03:20 PM (PDT)
Break
Oral
Tue Apr 06 03:20 PM -- 03:40 PM (PDT)
TensorFlow Lite Micro: Embedded Machine Learning for TinyML Systems
[
Paper PDF]
Oral
Tue Apr 06 03:40 PM -- 04:00 PM (PDT)
Scaling Distributed Training with Adaptive Summation
[
Paper PDF]
Oral
Tue Apr 06 04:00 PM -- 04:20 PM (PDT)
PipeMare: Asynchronous Pipeline Parallel DNN Training
[
Paper PDF]
Oral
Tue Apr 06 04:20 PM -- 04:40 PM (PDT)
EXPLORING THE LIMITS OF CONCURRENCY IN ML TRAINING ON GOOGLE TPUS
[
Paper PDF]
Oral
Tue Apr 06 04:40 PM -- 05:00 PM (PDT)
TT-Rec: Tensor Train Compression for Deep Learning Recommendation Models
[
Paper PDF]
Poster
Tue Apr 06 05:00 PM (PDT) @ Virtual
MicroNets: Neural Network Architectures for Deploying TinyML Applications on Commodity Microcontrollers
Poster
Tue Apr 06 05:00 PM (PDT)
Pufferfish: Communication-efficient Models At No Extra Cost
[
Paper PDF]
Poster
Tue Apr 06 05:00 PM (PDT)
A Deep Learning Based Cost Model for Automatic Code Optimization
[
Paper PDF]
Poster
Tue Apr 06 05:00 PM (PDT) @ Virtual
Cortex: A Compiler for Recursive Deep Learning Models
Poster
Tue Apr 06 05:00 PM (PDT) @ Virtual
In-network Aggregation for Shared Machine Learning Clusters
Poster
Tue Apr 06 05:00 PM (PDT)
EXPLORING THE LIMITS OF CONCURRENCY IN ML TRAINING ON GOOGLE TPUS
[
Paper PDF]
Poster
Tue Apr 06 05:00 PM (PDT)
TensorFlow Lite Micro: Embedded Machine Learning for TinyML Systems
[
Paper PDF]
Poster
Tue Apr 06 05:00 PM (PDT)
ModularNAS: Towards Modularized and Reusable Neural Architecture Search
[
Paper PDF]
Poster
Tue Apr 06 05:00 PM (PDT)
Data Movement Is All You Need: A Case Study on Optimizing Transformers
[
Paper PDF]
Poster
Tue Apr 06 05:00 PM (PDT) @ Virtual
TT-Rec: Tensor Train Compression for Deep Learning Recommendation Models
Poster
Tue Apr 06 05:00 PM (PDT)
Learning on Distributed Traces for Data Center Storage Systems
[
Paper PDF]
Poster
Tue Apr 06 05:00 PM (PDT)
Characterizing and Taming Model Instability Across Edge Devices
[
Paper PDF]
Session
Tue Apr 06 05:00 PM (PDT)
Poster Session 1
Invited Talk
Wed Apr 07 08:00 AM -- 08:50 AM (PDT)
Trustworthy AI
Break
Wed Apr 07 08:50 AM -- 09:10 AM (PDT)
Break
Oral
Wed Apr 07 09:10 AM -- 09:30 AM (PDT)
An Efficient Statistical-based Gradient Compression Technique for Distributed Training Systems
[
Paper PDF]
Oral
Wed Apr 07 09:30 AM -- 09:50 AM (PDT)
Adaptive Gradient Communication via Critical Learning Regime Identification
[
Paper PDF]
Oral
Wed Apr 07 10:10 AM -- 10:30 AM (PDT)
Rethinking Floating Point Overheads for Mixed Precision DNN Accelerators
[
Paper PDF]
Oral
Wed Apr 07 10:30 AM -- 10:50 AM (PDT)
Bit Error Robustness for Energy-Efficient DNN Accelerators
[
Paper PDF]
Break
Wed Apr 07 10:50 AM -- 11:10 AM (PDT)
Break - Visit the
Oral
Wed Apr 07 11:10 AM -- 11:30 AM (PDT)
RL-Scope: Cross-stack Profiling for Deep Reinforcement Learning Workloads
[
Paper PDF]
Oral
Wed Apr 07 11:30 AM -- 11:50 AM (PDT)
A Learned Performance Model for Tensor Processing Units
[
Paper PDF]
Oral
Wed Apr 07 11:50 AM -- 12:10 PM (PDT)
Accounting for Variance in Machine Learning Benchmarks
[
Paper PDF]
Oral
Wed Apr 07 12:10 PM -- 12:30 PM (PDT)
Larq Compute Engine: Design, Benchmark and Deploy State-of-the-Art Binarized Neural Networks
[
Paper PDF]
Break
Wed Apr 07 12:30 PM -- 01:30 PM (PDT)
Lunch Break / Visit the
Oral
Wed Apr 07 01:30 PM -- 01:50 PM (PDT)
IOS: Inter-Operator Scheduler for CNN Acceleration
[
Paper PDF]
Oral
Wed Apr 07 01:50 PM -- 02:10 PM (PDT)
Value Learning for Throughput Optimization of Deep Learning Workloads
[
Paper PDF]
Oral
Wed Apr 07 02:10 PM -- 02:30 PM (PDT)
ByzShield: An Efficient and Robust System for Distributed Training
[
Paper PDF]
Oral
Wed Apr 07 02:30 PM -- 02:50 PM (PDT)
FirePlace: Placing Firecraker Virtual Machines with Hindsight Imitation
[
Paper PDF]
Break
Wed Apr 07 02:50 PM -- 03:20 PM (PDT)
Break - Visit the
Oral
Wed Apr 07 03:20 PM -- 03:40 PM (PDT)
Nimble: Efficiently Compiling Dynamic Neural Networks for Model Inference
[
Paper PDF]
Oral
Wed Apr 07 03:40 PM -- 04:00 PM (PDT)
MicroRec: Efficient Recommendation Inference by Hardware and Data Structure Solutions
[
Paper PDF]
Oral
Wed Apr 07 04:00 PM -- 04:20 PM (PDT)
VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference
[
Paper PDF]
Oral
Wed Apr 07 04:20 PM -- 04:40 PM (PDT)
Accelerate Inference of CNNs for Video Analysis While Preserving Exactness Exploiting Activation Sparsity
[
Paper PDF]
Oral
Wed Apr 07 04:40 PM -- 05:00 PM (PDT)
sensAI: ConvNets Decomposition via Class Parallelism for Fast Inference on Live Data
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT) @ Virtual
IOS: Inter-Operator Scheduler for CNN Acceleration
Poster
Wed Apr 07 05:00 PM (PDT)
Rethinking Floating Point Overheads for Mixed Precision DNN Accelerators
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT) @ Virtual
Don't Forget to Sign the Gradients!
Poster
Wed Apr 07 05:00 PM (PDT)
A Learned Performance Model for Tensor Processing Units
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
Accounting for Variance in Machine Learning Benchmarks
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
Adaptive Gradient Communication via Critical Learning Regime Identification
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
Value Learning for Throughput Optimization of Deep Learning Workloads
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT) @ Virtual
VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference
Poster
Wed Apr 07 05:00 PM (PDT)
RL-Scope: Cross-stack Profiling for Deep Reinforcement Learning Workloads
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
MicroRec: Efficient Recommendation Inference by Hardware and Data Structure Solutions
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
Larq Compute Engine: Design, Benchmark and Deploy State-of-the-Art Binarized Neural Networks
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
Accelerate Inference of CNNs for Video Analysis While Preserving Exactness Exploiting Activation Sparsity
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT) @ Virtual
Bit Error Robustness for Energy-Efficient DNN Accelerators
Poster
Wed Apr 07 05:00 PM (PDT)
ByzShield: An Efficient and Robust System for Distributed Training
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
An Efficient Statistical-based Gradient Compression Technique for Distributed Training Systems
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
Nimble: Efficiently Compiling Dynamic Neural Networks for Model Inference
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
FirePlace: Placing Firecraker Virtual Machines with Hindsight Imitation
[
Paper PDF]
Poster
Wed Apr 07 05:00 PM (PDT)
sensAI: ConvNets Decomposition via Class Parallelism for Fast Inference on Live Data
[
Paper PDF]
Session
Wed Apr 07 05:00 PM (PDT)
Poster Session 2
Invited Talk
Thu Apr 08 08:00 AM -- 08:50 AM (PDT)
Machine Learning in Science: Applications, Algorithms and Architectures
[
Slides]
Break
Thu Apr 08 08:50 AM -- 09:10 AM (PDT)
Break - Visit the
Oral
Thu Apr 08 09:10 AM -- 09:30 AM (PDT)
Boveda: Building an On-Chip Deep Learning Memory Hierarchy Brick by Brick
[
Paper PDF]
Oral
Thu Apr 08 09:30 AM -- 09:50 AM (PDT)
Horizontally Fused Training Array: An Effective Hardware Utilization Squeezer for Training Novel Deep Learning Models
[
Paper PDF]
Oral
Thu Apr 08 09:50 AM -- 10:10 AM (PDT)
A Distributed Graph-Theoretic Framework for Automatic Parallelization in Multi-core Systems
[
Paper PDF]
Oral
Thu Apr 08 10:10 AM -- 10:30 AM (PDT)
Accelerating SLIDE Deep Learning on Modern CPUs: Vectorization, Quantizations, Memory Optimizations, and More
[
Paper PDF]
Oral
Thu Apr 08 10:30 AM -- 10:50 AM (PDT)
Scaling Polyhedral Neural Network Verification on GPUs
[
Paper PDF]
Break
Thu Apr 08 10:50 AM -- 11:10 AM (PDT)
Break - Visit the
Oral
Thu Apr 08 11:10 AM -- 11:30 AM (PDT)
SUOD: Accelerating Large-Scale Unsupervised Heterogeneous Outlier Detection
[
Paper PDF]
Oral
Thu Apr 08 11:30 AM -- 11:50 AM (PDT)
Lost in Pruning: The Effects of Pruning Neural Networks beyond Test Accuracy
[
Paper PDF]
Oral
Thu Apr 08 11:50 AM -- 12:10 PM (PDT)
Equality Saturation for Tensor Graph Superoptimization
[
Paper PDF]
Oral
Thu Apr 08 12:10 PM -- 12:30 PM (PDT)
Doping: A technique for Extreme Compression of LSTM Models using Sparse Structured Additive Matrices
[
Paper PDF]
Break
Thu Apr 08 12:30 PM -- 01:30 PM (PDT)
Lunch Break / Visit the
Oral
Thu Apr 08 01:30 PM -- 01:50 PM (PDT)
Swift for TensorFlow: A portable, flexible platform for deep learning
[
Paper PDF]
Oral
Thu Apr 08 01:50 PM -- 02:10 PM (PDT)
Amazon SageMaker Debugger: A System for Real-Time Insights into Machine Learning Model Training
[
Paper PDF]
Oral
Thu Apr 08 02:10 PM -- 02:30 PM (PDT)
FLAML: A Fast and Lightweight AutoML Library
[
Paper PDF]
Oral
Thu Apr 08 02:30 PM -- 02:50 PM (PDT)
To Bridge Neural Network Design and Real-World Performance: A Behaviour Study for Neural Networks
[
Paper PDF]
Break
Thu Apr 08 02:50 PM -- 03:20 PM (PDT)
Break - Visit the
Oral
Thu Apr 08 03:20 PM -- 03:40 PM (PDT)
Towards Scalable Distributed Training of Deep Learning on Public Cloud Clusters
[
Paper PDF]
Oral
Thu Apr 08 03:40 PM -- 04:00 PM (PDT)
Understanding and Improving Failure Tolerant Training for Deep Learning Recommendation with Partial Recovery
[
Paper PDF]
Oral
Thu Apr 08 04:00 PM -- 04:20 PM (PDT)
Wavelet: Efficient DNN Training with Tick-Tock Scheduling
[
Paper PDF]
Oral
Thu Apr 08 04:20 PM -- 04:40 PM (PDT)
Pipelined Backpropagation at Scale: Training Large Models without Batches
[
Paper PDF]
Remarks
Thu Apr 08 04:40 PM -- 05:00 PM (PDT)
Closing Remarks
Poster
Thu Apr 08 05:00 PM (PDT)
To Bridge Neural Network Design and Real-World Performance: A Behaviour Study for Neural Networks
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT) @ Virtual
Scaling Polyhedral Neural Network Verification on GPUs
Poster
Thu Apr 08 05:00 PM (PDT)
A Distributed Graph-Theoretic Framework for Automatic Parallelization in Multi-core Systems
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT)
Boveda: Building an On-Chip Deep Learning Memory Hierarchy Brick by Brick
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT) @ Virtual
Pipelined Backpropagation at Scale: Training Large Models without Batches
Poster
Thu Apr 08 05:00 PM (PDT)
Towards Scalable Distributed Training of Deep Learning on Public Cloud Clusters
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT)
Doping: A technique for Extreme Compression of LSTM Models using Sparse Structured Additive Matrices
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT)
Wavelet: Efficient DNN Training with Tick-Tock Scheduling
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT)
Swift for TensorFlow: A portable, flexible platform for deep learning
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT)
Equality Saturation for Tensor Graph Superoptimization
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT)
Horizontally Fused Training Array: An Effective Hardware Utilization Squeezer for Training Novel Deep Learning Models
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT) @ Virtual
Lost in Pruning: The Effects of Pruning Neural Networks beyond Test Accuracy
Poster
Thu Apr 08 05:00 PM (PDT)
Accelerating SLIDE Deep Learning on Modern CPUs: Vectorization, Quantizations, Memory Optimizations, and More
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT)
Amazon SageMaker Debugger: A System for Real-Time Insights into Machine Learning Model Training
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT)
SUOD: Accelerating Large-Scale Unsupervised Heterogeneous Outlier Detection
[
Paper PDF]
Poster
Thu Apr 08 05:00 PM (PDT) @ Virtual
Understanding and Improving Failure Tolerant Training for Deep Learning Recommendation with Partial Recovery
Session
Thu Apr 08 05:00 PM (PDT)
Poster Session 3
Workshop
Fri Apr 09 06:15 AM -- 03:00 PM (PDT)
Personalized Recommendation Systems and Algorithms
Workshop
Fri Apr 09 07:00 AM -- 04:00 PM (PDT)
Workshop of Graph Neural Networks and Systems (GNNSys'21)
Workshop
Fri Apr 09 07:00 AM -- 03:00 PM (PDT)
2nd On-Device Intelligence Workshop
Workshop
Fri Apr 09 07:45 AM -- 04:00 PM (PDT)
SysML4Health: Scalable Systems for ML-driven Analytics in Healthcare
Workshop
Fri Apr 09 08:00 AM -- 03:00 PM (PDT)
Journal of Opportunities, Unexpected limitations, Retrospectives, Negative results, and Experiences
Workshop
Fri Apr 09 08:00 AM -- 05:00 PM (PDT)
Benchmarking Machine Learning Workloads on Emerging Hardware