Mahesh Subramaniam, Michal Styszynski, and Himanshu Tambakuwala dive into the infrastructure side of AI — specifically, how data center networks need to be designed and built to handle the demands of large-scale AI workloads. This is a specialized, technically dense resource for network and infrastructure engineers.
Availability:
In StockBehind every large AI model is a data center network built to move enormous amounts of data reliably and fast, and AI Data Center Network Design and Technologies is focused entirely on that infrastructure layer. The three authors bring deep technical backgrounds in networking and infrastructure, and the book reflects that expertise throughout its 384 pages, covering the specific design considerations, technologies, and architectural decisions involved in building networks capable of supporting AI workloads at scale. This isn’t a book about AI models themselves, but about the physical and architectural foundation that makes training and running them possible — topics like network topology, bandwidth requirements, and the unique demands AI workloads place on infrastructure compared to traditional computing. It’s a genuinely specialized text, aimed at network engineers, infrastructure architects, and data center professionals who need to understand these requirements at a technical level rather than a conceptual one. For readers working at the infrastructure layer of the AI industry, this addresses a side of the field that gets far less coverage than the models themselves.
Authors: Mahesh Subramaniam, Michal Styszynski, Himanshu Tambakuwala.
Reviews
There are no reviews yet.