Chapter 2: Industry background
Abstract
This chapter focuses on the AI industry background. Firstly, the current status of popular neural network processors is introduced, including CPU, GPU, TPU, FPGA, and Huawei Ascend AI processor. Then, it is explained how CPU and GPU accelerate neural network. In the next part, the popular deep learning software frameworks are introduced, including Caffe, Tensorflow, Pytorch, and Huawei’s MindSpore. At the end of the chapter, the idea of intermediate representation (IR) and deep learning compilation framework is illustrated, which helps migrate algorithms between different frameworks. As an example, tensor virtual machine (TVM) is explained in detail.
Keywords
Huawei ascend AI processor; CPU; GPU; ...
Get Ascend AI Processor Architecture and Programming now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
