FlexFlow: A Flexible Dataflow Accelerator Architecture for Convolutional Neural Networks

By Wenyan Lu, Guihai Yan, Jiajun Li, Shijun Gong, Yinhe Han, Xiaowei Li. International Symposium on High-Performance Computer Architecture, 2017

324 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/ce6403e99465e5e8a48d5c2017fc23976e29fe59

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Electrical and Electronic Engineering

Citations in the corpus

Machine Learning at Facebook: Understanding Inference at the Edge (2019, background): “Recent work on hardware for machine learning and efficient training and inference has substantially advanced the state-of-the-art [41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77].”
Timeloop: A Systematic Approach to DNN Accelerator Evaluation (2019, background): “Dataflows have been shown to have a significant impact on performance and energy efficiency due to the resulting differences in data reuse and hardware utilization [6], [19], [22], [25], [31].”
Understanding Reuse, Performance, and Hardware Cost of DNN Dataflow: A Data-Centric Approach (2018, methodology): “C-brain [43], Flexflow [25], and analyzed the costbenefit tradeoff of three dataflows and explored the opportunity of adaptive dataflow based on the tradeoff.”
Domain-specific Genetic Algorithm for Multi-tenant DNNAccelerator Scheduling (2021, background): “While there has been significant prior work on scheduling a single DNN model efficiently over one (or more distributed) accelerators [65, 95, 100, 102, 115, 122], scheduling multiple models simultaneously is relatively unexplored, and is the focus of this work.”
vPipe: A Virtualized Acceleration System for Achieving Efficient and Scalable Pipeline Parallel DNN Training (2022, background): “Hybrid parallelism systems, including OWT [26], FlexFlow [30], etc, are designed to improve the training efﬁciency of such heterogenous DNNs. Speciﬁcally, these systems apply data parallelism to convolution layers and apply model parallelism to fully connected layers.”
FPGA-Based Accelerators of Deep Learning Networks for Learning and Classification: A Review (2019, background): “Processing Element (PE) Architecture in; (a) FlexFlow, (b) 2D-Mapping [118].”
Efficient Acceleration of Deep Learning Inference on Resource-Constrained Edge Devices: A Review (2023, background): “FlexFlow [311] is a scalable accelerator that exploits the parallelism of computing platforms and CNN architectures to support a flexible data flow and data reuse.”
AccPar: Tensor Partitioning for Heterogeneous Deep Learning Accelerators (2020, background): “Due to the high demand of computing power in DL applications, we have recently witnessed a phenomenal trend in which the landscape of computing has shifted from general-purpose processors to domain-specific architectures [16–110].”
MAERI: Enabling Flexible Dataflow Mapping over DNN Accelerators via Reconfigurable Interconnects (2018, background): “We demonstrate how Maeri is not only capable of running CONV, LSTM, POOL and FC layers, but also supports cross-layer mapping and sparsity.”