I was excited to get to review this book. After having read and reviewed Green mobile devices and networks [1], this book had me quite interested. I was not disappointed.

This book is about managing the energy used for computing systems and installations. As the subtitle states, the book discusses large-scale computing, not the level of sensors or small devices. There is nothing shallow in this book. However, a general caveat: the audience for this book is quite small; it is not for just anyone or for the average programmer or even manager. Also, the individual chapters of the book--each written by a different team--cover quite a broad range of issues. I doubt there are many people who will find equal interest in all of them. Fortunately, they all seem to be well written--the editor clearly deserves credit for a job well done.

The book is divided into nine chapters. Each covers a quite different aspect. All provide metrics and tools for application.

Chapter 1 discusses IBM’s Blue Gene project and how it approached the issue of energy efficiency. Any electrical engineer dealing with serious computing design can learn lessons from this chapter.

Chapter 2 is a quite unique view: how to implement objectives of energy efficiency in a compiler. I happily admit that the authors surprised me with some interesting ideas. Chapters 3 and 4 are similar, but rather than compilers, runtime systems are discussed, with 4 talking about multithreading.

Chapter 5 discusses how these issues may impact system reliability, with emphasis on storage systems. Most large-scale computing facilities today have massive storage, and the readership of this chapter may be quite larger than expected.

Chapters 6 and 7 discuss specialized systems/issues and their power management.

Finally, chapters 8 and 9 discuss computing centers. An idea in chapter 9 strongly caught my eye: in many data centers, the cost of electricity has already passed that of information technology (IT) equipment, and in the very near future this will become the norm for major computing environments.

If I may add my two cents, computer engineers need to start thinking less about how to rid systems of their heat and begin to think more about how to harvest the energy being given off of these systems--lessons I learned from the previously mentioned book about mobile devices [1]. From this book, one learns that the energy issue is far from limited to mobile devices, and any good computing manager needs to look much deeper into the future of the systems under his/her control.

Another word about the readership of this book. Generally, the readership may be limited to chief information officers (CIOs) and computer engineers. In some cases, however, the readership might also include heads of security. I am familiar with algorithms that read power graphs coming off motors to predict their future reliability. From my reading of this book, I think--I do not yet have proof of this--that power data from major computing centers may be a major concern for the security of many systems. Probably not that of a hotel chain, but quite likely that of a military or intelligence center. Better to be safe than surprised.

The book is well written and reads well. I admire it and recommend it for those readers who have these specific needs. It may prove useful or applicable in certain environments. I doubt its value for general coursework, though for certain electrical engineering and computer science areas I may be wrong. It is not trivial. I think it may prove more confusing to students than helpful and instructive. Use these methods for large, critical computing installations, for long-term integrity.