Customer Reviews

Computer System Architecture (3rd Edition)

5 star:		(4)
4 star:		(5)
3 star:		(7)
2 star:		(5)
1 star:		(3)

 

 

It's hard for me to believe this book didn't get good reviews. It's an awesome book that succintly details the design of a CPU. The author uses a hands on approach by showing you how to design a simple CPU from scratch. Even though simple, this CPU has a useable set of instructions (including IO ones), interrupts, and a memory subsytem. RTL is used to convey what is done at each instruction cyle, which is great. For me, this is the first time
I see a practical application of RTL. The book could have been more interesting if it provides an implementation in VHDL, or Verilog, but i guess that's left as an excercise for the reader.

I refer to the latest edition of this text by Mano and Kime. I tried to teach Comp. Architecture to a senior level BS class (in CS) from this book. The problem is that this book attempts to do too much and it wanders all over the subject area, leaving the students lost since they just cannot grasp the "big picture". This book gets progressively worse with every new edition! I had to extensively edit the contents to prepare lecture notes which could be comprehended. Halfway through the semester, when I came to the part on "microprogrammed control", I gave up and switched to Tanenbaum (Structured Comp Organization) instead. I would appreciate it if anyone could tell me about a better book on this subject. All the ones that I have looked at seem to be terrible in one way or another.

I have three of M. Mano's books and this one is the one I've liked the most. I found the author's exposition of the material good on average. I'm using this book for self-study. The book provides a nice overview of computer architecture by focusing on the basic concepts in manner that is not dependent on a particular real-life architecture.

I must say however that I gotten up to chapter 11 and have become totally unmotivated to finish the book (only two chapters left). I've gone through most of the interesting problems of each chapter but have no clue to the 'correctness' of some of them since the book doesn't come with solutions.

One thing that I really dislike about a book is the lack of an errata sheet. All of Mano's books I own lack an errata sheet. I've noticed some errors in the book although nothing major.

Recommendation: As far as I know there are no really good books in the area of computer architecture. If you're looking for a simple introductory book on computer architecture, I would recommend this book. Avoid this book if you're looking for a more technical treatment on the subject.

The book gives a detail introduction to computer hardware but it does not have some basic essentials: 1. No answers to the end of chapter problems were given. 2. No examples to problems with full solutions 3. No end of chapter reviews or glossary at end of text.

A study guide would be nice too.

As a student with an perfect record (GPA:4.0)with but a single class to go until graduation, I felt so uncomfortable with this text that I dropped the course. The meandering and lack of technical elaboration was frustrating and confusing. I felt it a shame that any University would approve such garbage. It is enough to make one curious about whether money is being exchanged under the table between the book publishers and schools.

Mano's book served as the main text for 2 courses I've taken this year (I'm a 2nd year Computer Science student). The book is quite informative. The chapter dealing with the basic computer (chap. 5) and the one dealing with assembly for this computer (chap. 6) are especially good. These chapters are full of details, and let you understand how a real computer works. The chapter about microprogramming is also good. The chapters about the CPU, I/O, computer memory and pipeline are less detailed, but give you a general feeling of how things work. Generally speaking, the book is very good, although it is not perfect; mistakes do pop up here and there, although they are not major. I hope a new edition will come out soon, because some of the material seems to be outdated (the estimated computation times used, for example, are just too slow -- one joked calling them mano seconds instead of nano seconds :)).

I found this book to be one of the most clearly written texts on the subject of computer architecture(as opposed to the terse, cryptic Astrom and Wittenmark books for control systems, for example.)

Now, a professor can easily test around the basics of any text book by surprising students with unexpected special cases, exceptions, and boundary conditions. But, that is not the fault of the text book itself. If any of those tricks actually appeared in a text book, they would be ignored and he/she would find new ones to test.

Noone can memorize how to think; knowing how to solve the worked examples in a textbook is only the beginning. That's why there are homework problems.

Unpopular professor tricks aside, I think this is a great book on the topic and it gave me the skills to perform certain assignments on my own during my first engineering job...skills that my employer thought he would have to teach to all us newbies. I was glad I had this book in school.

This book is not very hard to follow if you dive down into it. Everything presented is presented in an easy to understand, clear way. The exercises and hypothetical computer architecture gets a little annoying, why learn a language that doesn't exist when there are hundreds out there that could actually be used for something other than a course in hardware?

The problems are several. First of all, the author leaves many many avenues of obvious questions unexplored. This is a nice introduction to hardware, but it really doesn't get into much detail on some of the more complex areas such as ALU design. In keeping with his "skimming the surface" paradigm, the exercises at the end of the chapter are trivial to say the least. They will not challenge you if you were even remotely attentive while reading.

The book isn't the worst out there, but I'm sure there are better.

Let me start off by saying that I consider myself to have an excellent knowledge of computers (including hardware, software, and software design) and a reasonable knowledge of hardware design and electronic circuits. Despite this fact, every time I finish reading a section of this book, I'm completely lost.
You ever hear the expression "this reads like stereo instructions"? Well I think the author of that expression first said that after reading this book.
I'm not an idiot, but this book is just written in a style that is way too difficult to understand. Sometimes I'm reading about a topic that I already know everything about, but I still can't understand what the book is talking about. I have to read and reread each sentence till I can finally safely conclude that "this" is what he's trying to say. However, if you have no knowledge of the topic prior to reading it, that is an impossible task.
So far, after reading each topic from this book (and being completely confused when I'm done), I've had to look the topic up on the internet, and within minutes I completely understand it. The topics aren't hard. They're almost easy, but his writing style is so difficult to understand that you feel like you're reading about nuclear physics.
If your a student and you have to read this book, I strongly reccommend that you don't get discouraged, and just turn to the internet to understand the material.

The book isn't really all that bad, the problem is that it can be confusing to some readers. Instead of giving a list of problems, I am going to go into some detail about one specific thing wrong with the book, which could be a BIG problem for some readers. It is this:

It seems like once or twice every couple paragraphs, Mano will give a one-sentence explanation or description of something; and then a sentence or two later in the same paragraph, he will attempt to make the same point again but by rewording and rearranging the original sentence. I found this so incredibly distracting that I could not finish the book. I think Mano is doing this, i.e. repeating himself, to embed a concept into the student's mind. Not a bad idea, but it is poorly implemented by Mano...

what was so distracting about this, is that even if the reader feels he understands the concept the first time it is presented, when he then reads the reworded version, the tendency is to say "hey, wait a minute, is this the same thing he said before, or do the two statements conflict? or is it new information? Or did I not understand the first time??" So then I, at least, would go back and reread the first statement, then again compare it to the second, until I was able to answer these questions for myself. So it really slowed down my reading. I don't know, maybe it's just the way I read.

I should mention that I was mainly trying to learn from the book by self-study, i.e. no professor, although I did take a microprocessors course about 15 years ago.

Well, I guess I was verbose :-> but I wanted to inform so that anyone who thinks that they also will be distracted by this, buy a different book (like Rafiquzzaman)