Jakub Szefer's New Book Takes on Secure Processor Design

Recent years have seen an increased interest in exploring and designing hardware features inside computer processors that protect confidentiality and the integrity of the code and data executing on the processor. And with that increased need, Jakub Szefer’s new book fills a need. 

“Principles of Secure Processor Architecture Design” presents the different challenges of a relatively new field. It educates readers on how different challenges have been solved in the past and the principles for the design of new secure processor architectures. The book also presents existing research challenges and potential new research directions. Additionally, it presents numerous design suggestions, and discusses pitfalls and fallacies that designers should avoid. 

“It’s one of those things that people knew there should be such a book and that somebody should write it, but the question was, who’s going to do it?” said Szefer, assistant professor of electrical engineering & computer science. “I think this will be a missing piece, which will be helpful to fill in.”

Szefer said first- and second-year graduate students are the ideal audience for the book. “But I think it’s also good for industry, for instance, if someone is a computer architect but they’re trying to get into security, maybe they can use it as a way to learn what the architecture topics that are related to security that they should know about.” 

Traditional processor architecture research focuses on performance, efficiency, and energy as first-order design objectives, while secure processor architecture design has security as the first-order design objective (although the traditional design aspects are also considered). 

Szefer said the field of research emerged about 15 years ago, but interest has increased significantly in recent years. 

“It became more interesting to people with the emergence of cloud computing, now that you can remotely access all different types of servers and are more aware of the hardware they’re using,” he said.

Szefer worked on the book for about two years. In addition to keeping up-to-date with current research, writing the book meant keeping up with the headlines. Near the book’s completion, news came out about vulnerabilities in Intel chips, requiring Szefer to add some related information. With no sign that the field is slowing down, Szefer said he’ll likely update the book every few years.