Nexus: a GPU cluster engine for accelerating DNN-based video analysis
Explore this paper's citation graph
Summary
Nexus is a fully implemented system that includes cluster-scale resource management that performs detailed scheduling of GPUs, reasoning about groups of DNN invocations that need to be co-scheduled, and moving from the conventional whole-DNN execution model to executing fragments ofDNNs.
- Type
- article
- Published
- 2019-10-27
- Cited by
- 350
- References
- 39
- Access
- Open access
- OpenAlex
- https://openalex.org/W2982157693
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:204812163
Keywords
Computer science, Latency (audio), Nexus (standard), Software deployment, Scheduling (production processes)
References
- Combinatorial Optimization: Theory and Algorithms
- A large-scale car dataset for fine-grained categorization and verification
- CNN Features Off-the-Shelf: An Astounding Baseline for Recognition
- Sparrow: distributed, low latency scheduling
- Gradient-based learning applied to document recognition
- Exploiting sparseness in deep neural networks for large vocabulary speech recognition
- Omega: flexible, scalable schedulers for large compute clusters
- Caffe: Convolutional Architecture for Fast Feature Embedding
- Two-Stream Convolutional Networks for Action Recognition in Videos
- Learning and Transferring Mid-level Image Representations Using Convolutional Neural Networks
- Mesos: A Platform for Fine-Grained Resource Sharing in the Data Center
- Deep Residual Learning for Image Recognition
- Deep Face Recognition
- TensorFlow: a system for large-scale machine learning
- DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
- MCDNN: An Approximation-Based Execution Framework for Deep Stream Processing Under Resource Constraints
- Fast Video Classification via Adaptive Cascading of Deep Models
- Fast Cholesky factorization on GPUs for batch and native modes in MAGMA
- Clipper: A Low-Latency Online Prediction Serving System
- Live Video Analytics at Scale with Approximation and Delay-Tolerance
Cited by
- A High-Accuracy Implementation for Softmax Layer in Deep Neural Networks
- Towards GPU Utilization Prediction for Cloud Deep Learning
- Serving DNNs like Clockwork: Performance Predictability from the Bottom Up
- Enabling Cost-Effective, SLO-Aware Machine Learning Inference Serving on Public Cloud
- Incremental and Approximate Computations for Accelerating Deep CNN Inference
- MARBLE: A Multi-GPU Aware Job Scheduler for Deep Learning on HPC Systems
- Better Never Than Late: Timely Edge Video Analytics Over the Air
- Reducto: On-Camera Filtering for Resource-Efficient Real-Time Video Analytics
- An Overview of Efficient Interconnection Networks for Deep Neural Network Accelerators
- Accelerating Multi-Model Inference by Merging DNNs of Different Weights
- TurboTransformers: an efficient GPU serving system for transformer models
- InferLine: latency-aware provisioning and scaling for prediction serving pipelines
- PipeSwitch: Fast Pipelined Context Switching for Deep Learning Applications
- GSLICE: controlled spatial sharing of GPUs for a scalable inference platform
- Lazy Batching: An SLA-aware Batching System for Cloud Machine Learning Inference
- Srift: Swift and Thrift Cloud-Based Distributed Training
- Distream: scaling live video analytics with workload-adaptive distributed edge intelligence
- E2bird: Enhanced Elastic Batch for Improving Responsiveness and Throughput of Deep Learning Services
- Clownfish: Edge and Cloud Symbiosis for Video Stream Analytics
- ShadowVM: accelerating data plane for data analytics with bare metal CPUs and GPUs
Related papers
- Methods and Trends for Water-Energy-Carbon Emissions Nexus Assessment
- Review of water-energy-food Nexus tools to improve the Nexus modelling approach for integrated policy making
- Studying international doctoral researchers: nexus analysis as a mode of inquiry
- Ten principles to integrate the water-energy-land nexus with climate services for co-producing local and regional integrated assessments.
- Nexus approaches to global sustainable development
- The Food-Energy-Water Security Nexus: Definitions, Policies, and Methods in an Application to Malawi and Mozambique