Artificial Intelligence Hardware Design. Albert Chun-Chen Liu

      

      Table of Contents

      1  Cover

      2  Series Page

      3  Title Page

      4  Copyright Page

      5  Author Biographies

      6  Preface

      7  Acknowledgments

      8  Table of Figures

      9  1 Introduction 1.1 Development History 1.2 Neural Network Models 1.3 Neural Network Classification 1.4 Neural Network Framework 1.5 Neural Network Comparison Exercise References

      10  2 Deep Learning 2.1 Neural Network Layer 2.2 Deep Learning Challenges Exercise References

      11  3 Parallel Architecture 3.1 Intel Central Processing Unit (CPU) 3.2 NVIDIA Graphics Processing Unit (GPU) 3.3 NVIDIA Deep Learning Accelerator (NVDLA) 3.4 Google Tensor Processing Unit (TPU) 3.5 Microsoft Catapult Fabric Accelerator Exercise References

      12  4 Streaming Graph Theory 4.1 Blaize Graph Streaming Processor 4.2 Graphcore Intelligence Processing Unit Exercise References

      13  5 Convolution Optimization 5.1 Deep Convolutional Neural Network Accelerator 5.2 Eyeriss Accelerator Exercise References

      14  6 In‐Memory Computation 6.1 Neurocube Architecture 6.2 Tetris Accelerator 6.3 NeuroStream Accelerator Exercise References

      15  7 Near‐Memory Architecture 7.1 DaDianNao Supercomputer 7.2 Cnvlutin Accelerator Exercise References

      16  8 Network Sparsity 8.1 Energy Efficient Inference Engine (EIE) 8.2 Cambricon‐X Accelerator 8.3 SCNN Accelerator 8.4 SeerNet Accelerator Exercise References

      17  9 3D Neural Processing 9.1 3D Integrated Circuit Architecture 9.2 Power Distribution Network 9.3 3D Network Bridge 9.4 Power‐Saving Techniques Exercise References

      18  Appendix A: Neural Network Topology

      19  Index

      20  End User License Agreement

      List of Tables

      1 Chapter 1Table 1.1 Neural network framework.

      2 Chapter 2Table 2.1 AlexNet neural network model.

      3 Chapter 3Table 3.1 Intel Xeon family comparison.Table 3.2 NVIDIA GPU architecture comparison.Table 3.3 TPU v1 applications.Table 3.4 Tensor processing unit comparison.

      4 Chapter 5Table 5.1 Efficiency loss comparison.Table 5.2 DNN accelerator performance comparison.Table