Digital computers have found their way into every aspect of modern life. They are now essential for work and ubiquitous in leisure, and used from childhood to retirement age, through every stage of adult life. Although the number of digital computers in operation today is already mind boggling, it continues to grow at a fast pace. We seem to be at a loss as to where this ever-more digital life will lead. Some dream of a future where robots and artificial intelligence (AI) will do the heavy lifting, freeing humans for a life of creative work, while others fear intelligent machines spell doom for humanity. Whatever the outcome, one thing is certain: computing by digital machines is bound to leave its mark.
Remarkably, virtually all digital computers in use today have the same von Neumann architecture, which has proven highly effective for single computers and for vast clusters with thousands of nodes. It is easy to verify that the von Neumann architecture is the “winner” just by counting how many of those machines are sold today and contrasting that number to the count of machines of a different kind. But the interesting question is why is this the case. Were there alternative kinds of machines that were just not good enough? And, even more importantly, are there plausible competing alternatives today? Steiglitz addresses these two fundamental questions in The discrete charm of the machine, while at the same time providing a remarkably complete history of computing and of the machines that were invented to improve that process in one way or another. More precisely, the book discusses the advantages and disadvantages of analog and discrete computation, carefully building an argument from the ground up, going all the way to quantum physics. In the process, the book explains why digital computers are better; why data in digital form is easier to manipulate, transmit, and preserve; and how digitizing analog signals can be done without noticeable loss by our senses.
This is a truly remarkable book that every computer scientist should read. It is meticulously written, with crisp logic and rigorous arguments and yet very little math. Six key points are reinforced throughout the book, and discussed and revisited at different levels of granularity in a way that only the greatest teachers can articulate. The book also provides a glimpse of the two major discussion topics among computer scientists today: quantum computers and machine learning, with hints of what to realistically expect of them in the foreseeable future. It is also worth mentioning that the book touches on quite a bit of the thinking of the great minds that brought us to where we are today, with anecdotes about the life of the author as well. The result is a quite compelling and engaging narrative that humanizes the subject in a powerful way. In summary, Steiglitz treats us to a book that is not only thoroughly informative, but also very hard to put down.
More reviews about this item: Amazon, Goodreads