書籍：Efficient Processing of Deep Neural Networks

@Vengineerの戯言 : Twitter
SystemVerilogの世界へようこそ、すべては、SystemC v0.9公開から始まった

このツイートを見て、レポートを読んでみた。

Want to learn how to evaluate and compare the huge number of deep learning processors being developed? Check out our latest article for @SSCSociety Magazine on "How to Evaluate Deep Neural Network Processors: TOPS/W (Alone) Considered Harmful” https://t.co/pDNq6LLavU pic.twitter.com/7WbxLfzlUj
— EEMS Group @ MIT (PI: Vivienne Sze) (@eems_mit) 2020年9月2日

この本、「Efficient Processing of Deep Neural Networks」の宣伝っぽい。

www.morganclaypoolpublishers.com

Table of Contents

Part I Understanding Deep Neural Networks
1 Introduction
2 Overview of Deep Neural Networks

Part II Design of Hardware for Processing DNNs
3 Key Metrics and Design Objectives
4 Kernel Computation
5 Designing DNN Accelerators
6 Operation Mapping on Specialized Hardware

Part III Co-Design of DNN Hardware and Algorithms
7 Reducing Precision
8 Exploiting Sparsity
9 Designing Efficient DNN Models
10 Advanced Technologies
11 Conclusion

お値段