Nielsen Chuang Preskill Watrous

You have worked through a wiki article. You have watched a Watrous lecture. You have run Qiskit on a free IBM backend. And you are starting to feel a pull you did not expect: you want to sit with one text, in the same notation, for a hundred hours — the way a physics student sits with Resnick–Halliday–Walker or a mathematics student sits with Apostol. You want a book.

Quantum computing has exactly three books that qualify. Not ten, not five — three. The rest are supplements, companions, or niche references. The reason is historical: the field turned from "a few physicists' curiosity" into "a full academic subject" only in the late 1990s, which means there are not yet fifty textbooks competing for your attention. Three serious texts cover the subject; almost every professional learned from some combination of them.

This chapter is not a book review. It is a deployment guide: which book to open for which question, how they overlap and where they diverge, what each one is bad at, which you can read for free and which cost ₹7,500 on Amazon India, and — at the end — two worked reading plans that map the books onto real schedules. Work through it, pick one book to start with today, and the next six months of self-study has a spine.

The three books, at a glance

Before the detailed tour, a one-line sketch of each.

Nielsen and Chuang, Quantum Computation and Quantum Information (Cambridge University Press, 10th-anniversary edition 2010). Written by Michael Nielsen and Isaac Chuang, started in 1997, published in 2000, revised in 2010. 702 pages. Covers everything the field knew by 2000 at encyclopaedic depth. Costs around ₹7,500 in India for the paperback; the PDF is not legally free.

John Preskill's lecture notes on Quantum Computation at Caltech, course Ph229 — a 500-ish-page manuscript freely available at theory.caltech.edu/~preskill/ph229. Written starting 1997, revised continuously since. Free, forever, in PDF. Preskill teaches Ph229 every few years and the notes improve each round.

John Watrous, The Theory of Quantum Information (Cambridge University Press, 2018), 598 pages, free PDF at cs.uwaterloo.ca/~watrous/TQI. A graduate-level monograph on the mathematical structure of quantum information — channels, entropies, measurements, entanglement — with essentially no coverage of algorithms or hardware.

Why this triangular picture is not a ranking: Nielsen and Chuang is the "default" textbook, but a researcher working on quantum Shannon theory will reach for Watrous first; a student learning the surface code will reach for Preskill first. Different corners for different questions. The best outcome of this chapter is that you learn to aim at the right corner before opening anything.

Nielsen and Chuang — the encyclopaedia

Michael Nielsen (a Los Alamos then Perimeter then independent writer) and Isaac Chuang (MIT) began writing what became Quantum Computation and Quantum Information in 1997. The book they produced in 2000 was the first serious unified textbook on the field — before it, you had to read Preskill's notes, some chapters of Peres' Quantum Theory: Concepts and Methods, and a stack of papers. The 2010 tenth-anniversary edition is what you buy today; it has minor corrections but is otherwise identical.

Structure — 12 chapters, 4 parts

The book divides into four parts:

Part I — Fundamental Concepts (chapters 1–2). Introduction to the field, then a linear-algebra bootcamp followed by the postulates of quantum mechanics. Chapter 2 is about 80 pages and contains every linear-algebra tool the rest of the book uses — tensor products, spectral theorem, polar decomposition, density operators introduced here.

Part II — Quantum Computation (chapters 3–7). Classical computation recap, then the quantum circuit model, then algorithms (chapter 5 on QFT and phase estimation, chapter 6 on Grover, chapter 7 on Shor and hidden subgroup). This is the part most people read first and most often.

Part III — Quantum Information (chapters 8–12). Chapter 8 on quantum noise and channels. Chapter 9 on distance measures (fidelity, trace distance). Chapter 10 on quantum error correction — the textbook's QEC chapter, which for years was the standard treatment. Chapters 11 and 12 on classical Shannon theory and then quantum Shannon theory (entropies, channel capacities).

Appendices — a useful trio covering classical complexity, probability, and number theory. Read the number-theory appendix before reading chapter 7 on Shor — it is the shortest path into the modular arithmetic Shor relies on.

What N&C is strong at

• Encyclopaedic coverage. Almost any foundational question you have — "what is the exact definition of fidelity?", "how does the stabiliser formalism work?", "what does Shor's algorithm look like end to end?" — has a careful multi-paragraph answer somewhere in this book.
• Worked problems. Every chapter has 10–30 exercises with partial solutions in an online companion. Doing half of chapter 2's exercises leaves you with a working linear-algebra toolkit.
• A unified notation. The book fixes the notation that most of the field adopted — computational basis, circuit diagrams, density-operator conventions.

What N&C is weak at

• Age. Published 2000, revised 2010. No coverage of post-2010 developments: topological quantum codes beyond the basics, magic-state distillation, variational algorithms, NISQ, Google's supremacy experiments, any of the recent Shor's-algorithm engineering work. For state of the art, supplement with arXiv.
• Some chapters feel thin. Chapter 10 on QEC is excellent for the basics but stops before the surface code became central; chapter 12 on quantum Shannon theory is a sketch where Watrous gives a full treatment.
• Availability in India. Paperback at around ₹7,500 on Amazon.in; hardcover more. PDFs circulate but are not legal. A campus library with a Cambridge subscription is your best legal-free route.

Why you should still own it: even with the age issue, N&C is the book every working quantum-computing person has read, the reference everyone cites, and the shared vocabulary of the field. If you can afford one book, this is the one.

Which chapters to read, in what order

• First pass (first 2–3 months of study): chapters 1, 2, 4, 5. That is 300 pages. It gives you the linear-algebra foundations, the circuit model, and the core algorithms.
• Second pass: chapters 6, 7, 8, 10. Another 250 pages. Grover, Shor, noise, error correction — the remaining essentials.
• Third pass, optional: chapters 3, 9, 11, 12. Classical recap, distance measures, Shannon theory. Skip if you are not doing information theory; read carefully if you are.

Preskill's Ph229 notes — the free deep dive

John Preskill is a Caltech theorist who started teaching the quantum-computation course Ph229 in 1997 — the same year Nielsen and Chuang started writing their book. His lecture notes grew into a 500-ish-page manuscript that covers substantially the same material as N&C plus several topics N&C does not touch — and they are free, forever, at theory.caltech.edu/~preskill/ph229. If you have an internet connection and no budget, Preskill is your textbook.

Structure — chapters by topic, roughly

Preskill does not number his chapters like a textbook; they are chapter-sized PDFs named by topic. The core sequence:

• Chapter 1: overview and fundamental concepts. Covers the same ground as N&C chapter 1, slightly more conversational.
• Chapters 2–3: foundations of quantum theory (states, measurement, density operators, unitary evolution) and the formalism (tensor products, partial trace, channels introduced).
• Chapter 4: quantum entanglement. Bell inequalities, monogamy, a long treatment of bipartite entanglement measures.
• Chapter 5: classical and quantum circuits, reversible computing.
• Chapter 6: quantum algorithms — Deutsch, Simon, Grover, Shor.
• Chapter 7: quantum error correction. This is the chapter where Preskill pulls ahead of N&C — it is longer, covers the stabiliser formalism in more depth, and introduces topological ideas.
• Chapter 8: fault-tolerant quantum computation, magic-state distillation, threshold theorems. N&C only gestures at this; Preskill gives the full construction.
• Chapter 9: topological codes. The surface code, toric code, and the mathematics behind them. Essentially absent from N&C.
• Chapter 10: quantum Shannon theory. Entropies, channel capacities, coding theorems. A solid treatment, sharper in places than N&C chapter 12, less formal than Watrous.

What Preskill is strong at

• Prose. Preskill writes the way a senior researcher talks when they are being generous with their time. Sentences are longer than N&C's; the exposition is more relaxed; the motivation for each step is usually explicit. On hard topics like the threshold theorem or the topological-code geometry, he is often the clearest source in print.
• Topics N&C does not cover. Fault tolerance, magic-state distillation, topological codes, modern Shannon theory. If you are going anywhere near QEC research, Preskill's chapters 7–10 are the default reading.
• Free. You can download all nine chapters this afternoon. From anywhere. No library required.
• Updated continuously. Preskill edits the notes when the field moves. Mistakes are corrected within a few years of appearing; recent results occasionally get mentioned in footnotes.

What Preskill is weak at

• Inconsistent finish. The manuscript is lecture notes, not a polished textbook. Some chapters are camera-ready quality; others have unfinished sections, occasional typos, cross-references to exercises that do not appear. You tolerate rough edges in exchange for depth.
• Fewer exercises. N&C has hundreds of exercises with partial solutions; Preskill has dozens, more scattered.
• No index. Each chapter is a PDF on its own; there is no book-level index. Searching is "grep across nine PDFs" — Preview on a Mac does this; Adobe Reader's "search across PDFs" works on Windows.

Why Preskill is often the right answer despite being called "just notes": after chapter 2 on foundations, it is the clearest source on error correction and fault tolerance in existence. Researchers in those sub-fields often cite "Preskill's notes, chapter 7" as if it were a textbook, because functionally it is.

Watrous's TQI — the mathematical reference

John Watrous is a Waterloo-then-IBM researcher whose career has been spent on the mathematical foundations of quantum information. His 2018 book, The Theory of Quantum Information, published by Cambridge and also freely available as a PDF at cs.uwaterloo.ca/~watrous/TQI, is the definitive treatment of the formal structure of the subject: states, channels, measurements, entropies, entanglement, semidefinite programming bounds.

Structure — 8 chapters, pure mathematics

• Chapter 1: mathematical preliminaries — inner products, norms, operators, spectral theorem. A fast refresher that assumes you have had a first course in linear algebra over ℂ.
• Chapter 2: basic notions — states, measurements, channels. The definitions are introduced with mathematical precision; there is no hand-waving.
• Chapter 3: similarity and distance between states and channels. Fidelity, trace distance, diamond norm, induced norms. The diamond norm treatment is the field's canonical reference.
• Chapter 4: unital channels and majorisation. Channel decompositions, mixing times.
• Chapter 5: quantum entropy and source coding. Von Neumann entropy, quantum Shannon theory, compression.
• Chapter 6: bipartite entanglement. Measures, convertibility, LOCC.
• Chapter 7: permutation-invariance and exchangeable states. De Finetti theorems.
• Chapter 8: quantum channel capacities. The definitive treatment — multiple capacity notions, coding theorems, open problems.

What Watrous is strong at

• Mathematical rigour. Every theorem is stated precisely; every proof is given in full; every reference is to an original paper. If you want to know why a result is true rather than that it is true, this is the book that will tell you.
• Channels and entropies. Watrous's treatment of CPTP maps, the diamond norm, and channel capacities is the cleanest anywhere. Researchers in quantum Shannon theory treat this book as their standard reference.
• Semidefinite programming. The book develops SDP as a central tool in quantum information and uses it repeatedly. No other textbook takes SDP this seriously.

What Watrous is weak at

• No algorithms, no hardware. There is almost nothing about Shor's, Grover's, or Deutsch–Jozsa in Watrous; there is nothing at all about superconducting qubits, trapped ions, or any physical realisation. This is a mathematics book whose subject happens to be quantum information. If you want algorithms and engineering, you read N&C or Preskill.
• Prerequisites are heavy. A confident grasp of linear algebra over ℂ is assumed. Measure-theoretic instincts help. A first-year Indian undergraduate often finds Watrous's opening chapter more intimidating than N&C's opening chapter, even though they cover the same material, because Watrous assumes you will not be surprised by a tensor product appearing in the first ten pages.
• The prose is dense. Watrous is a clean writer, but his sentences are built for precision, not warmth. Reading him feels like reading a mathematics monograph — because that is what it is.

Why Watrous is still essential: once you are doing quantum information theory as research, Watrous is where the proofs live. For the first two years of your study you may never crack it open; in the third year you will start every library session by reaching for it.

Honorable mentions — the supplementary books

Beyond the triad, a handful of books are genuinely useful as supplements or alternatives for specific learning styles.

• N. David Mermin, Quantum Computer Science: An Introduction (Cambridge, 2007). ~240 pages. The slimmest serious introduction. Written for CS students who know some linear algebra but no quantum mechanics. Covers the algorithms cleanly, skips the physics. Better as a first book than N&C for many CS-background readers; thinner on QEC and channels.

• Wolfgang Scherer, Mathematics of Quantum Computing: An Introduction (Springer, 2019). ~750 pages. A textbook that reads like a workbook — every definition is followed by examples, every theorem by detailed proofs. Heavy on exercises. Useful if the Nielsen-and-Chuang style feels too dense and you want more hand-holding on the math.

• Noson Yanofsky and Mirco Mannucci, Quantum Computing for Computer Scientists (Cambridge, 2008). A CS-flavoured introduction, organised around algorithms and programming. Lighter on information theory, heavier on complexity. Good third book after a first serious treatment.

• Jack Hidary, Quantum Computing: An Applied Approach (Springer, 2nd ed 2021). ~500 pages. The most code-forward of the introductory books — every algorithm comes with Cirq and Qiskit implementations. Good companion to our SDKs guide.

• Phillip Kaye, Raymond Laflamme, and Michele Mosca, An Introduction to Quantum Computing (Oxford, 2007). A focused 300-page treatment aimed at CS undergraduates. Clean on Shor and Grover, light on QEC. Good middle option between Mermin and N&C.

• Classiq, PennyLane, and Qiskit learning portals. Not books, but the Qiskit Textbook (now partly folded into IBM Quantum Learning — see that chapter) and PennyLane's tutorial site fill the same role as an introductory textbook for many learners who prefer code-first exposition.

Indian-authored introductions

Indian contributions to quantum-computing pedagogy are newer but growing. A few to keep an eye on:

• Indian Institute of Technology lecture notes. IIT Madras (Prabha Mandayam), IIT Delhi, IISc Bangalore (Apoorva Patel), and IIT Bombay have published open lecture notes on quantum computing and quantum information — most available as PDFs on institute pages. The quality varies; the IIT Madras notes on quantum algorithms and IISc's notes on quantum information are both well-regarded.

• NPTEL courses. The National Programme on Technology Enhanced Learning hosts free video lectures from Indian faculty, several on quantum computing and quantum information. nptel.ac.in — search "quantum computing" — has courses by IIT faculty, many with accompanying notes in English and sometimes Hindi.

• Apoorva Patel and others, occasional chapters in Indian-published textbooks on advanced quantum mechanics that touch on quantum information.

• Hindi and regional language resources. These are still thin, but QWorld's India chapter has begun translating introductory material; YouTube channels like Varun Mayya's physics videos and CodeBasics touch on quantum computing in Hindi, though they are not textbook replacements.

Why Indian-authored resources matter even when imported books exist: language and cultural idiom shape how quickly a concept lands. An explanation of superposition that reaches for cricket and Diwali examples, written in Hindi or Tamil, reaches a reader the American textbooks never will. The resources are thin today; they will thicken over the next decade as the National Quantum Mission trains more faculty.

How to read these books from India

Most Indian students do not have a home copy of Nielsen and Chuang, and many do not have campus library access to Cambridge University Press e-books. Three practical routes:

1. Preskill and Watrous first. Both are legally free PDFs. If you can only read one of the three, and you have no budget, make it Preskill for a broad course, Watrous if your taste runs mathematical. Save N&C for later, when you can afford it or reach a library.

2. Buy N&C paperback. ₹7,500 on Amazon.in for a book you will use for a decade is not a bad investment if quantum computing is your field. Used copies sometimes appear on OLX and college bulletin boards at a discount.

3. Use a library. Most IIT, IISc, and central-university libraries have Cambridge e-book subscriptions; undergraduate readers can usually walk in and read. The British Council library in major cities often stocks the paperback. State-level technical universities increasingly have it as well.

Example: reading plan for a JEE student

The first worked example — a concrete plan for a 17-year-old who has just finished JEE Advanced and is heading into a CS or engineering-physics undergrad in the monsoon.

Example: reading plan for a second-year CS undergraduate

The second worked example — a 19-year-old who has a year of CS under their belt (linear algebra, discrete math, probability, some algorithms) and wants to move from curiosity to research.

Common confusions about the books

• "I should read Nielsen and Chuang cover to cover." Almost no one has. N&C is a reference, not a novel. Read chapters 1, 2, 4, 5 first (~300 pages); revisit specific later chapters when you need them. Treating it as a linear novel produces months of unproductive reading.

• "Preskill's notes are outdated because they are free." False. Preskill edits them continuously; many chapters are more current than N&C (chapter 9 on topological codes in particular has been revised in the last five years). The free price reflects academic generosity, not neglect.

• "Watrous is the best book and I should start there." Only if you are a mathematics student with a strong linear-algebra background and specific interest in information-theoretic foundations. For most readers — including most CS undergraduates — Watrous is the third book, not the first.

• "Mermin is 'only' for CS students." Mermin is fine for physics students too — it is a clean, narrow introduction that gets you to Shor and Grover with minimum fuss. Physics students sometimes prefer its pacing to N&C's encyclopaedia style.

• "Hidary or Yanofsky-Mannucci can replace N&C." They are good supplements, not replacements. The canonical reference is still N&C; the application-oriented books fill in the code-practice gap. Read one of them alongside N&C, not instead of it.

• "There must be a perfect Indian-authored introduction to quantum computing." There is not yet one book that plays the role N&C plays, written by an Indian author for Indian students. The IIT lecture notes, NPTEL courses, and forthcoming resources from the National Quantum Mission are the building blocks; the canonical book remains to be written. If you are a future academic reading this: that book is yours to write.

Going deeper

How to actually read a textbook chapter

Most students read textbooks the way they read novels — left to right, one page after another. This is the wrong technique for mathematics and quantum computing. The better method:

1. First pass — five minutes per chapter. Flip through. Read the chapter introduction, the section headings, every theorem statement, the summary. Do not read the proofs. Do not read the prose. The goal is a mental map — "this chapter contains four results, about these four objects."

2. Second pass — the prose and pictures. Read the chapter sequentially, slowly. Mark every equation you cannot derive on the spot. For each theorem, read the statement and the intuitive gloss; skim the proof.

3. Third pass — the proofs and exercises. Go back to the marked equations; derive them yourself on paper. Read the proofs carefully. Do half the exercises. This is where the learning happens; the first two passes were preparation.

4. Fourth pass, optional — the exercises you skipped. Only do this if you are committing to mastery. Most readers skip it; most pass with a working-but-not-deep understanding. Both are valid endpoints depending on your goals.

A chapter done this way takes 10–20 hours for a serious N&C chapter. This is why N&C's twelve chapters imply 150–250 hours of reading for a first pass. Plan your months accordingly.

Exercises — why they are not optional

The biggest difference between people who learn quantum computing and people who do not finish is exercise discipline. The exercises in N&C and Preskill are not busywork — they are where the concepts are operationalised. Working through exercise 2.61 (Schmidt decomposition on a specific state) is what converts "I know what Schmidt decomposition is" into "I can compute Schmidt decompositions."

A reasonable rule: for every 10 pages of reading, do 3–5 exercises. Do not skip the computational ones; they build the numerical fluency that makes research feasible.

When to read the original papers instead

At some point — usually around your second year of serious study — textbooks stop being the leading edge of what you need. Textbooks are always 5–10 years behind the field because they have to be written, edited, and typeset. The remedy is arXiv. A rough heuristic:

• Fundamentals (linear algebra, postulates, circuit model, Shor, Grover, basic QEC): read textbooks.
• Current research (specific new algorithms, recent hardware results, modern error-correction techniques, anything from the last 5 years): read papers from arXiv.

The transition is gradual. At month 6 you might read 10 pages of textbook for every 2 pages of arXiv; at month 24 those numbers might flip. See the arXiv reading chapter for the habit-building.

The case for ordering the books differently

Standard advice is N&C first, Preskill second, Watrous third. Two sensible alternatives:

• Preskill first if you have no budget. Preskill is free and covers the same core material as N&C for chapters 1–6; start there if you cannot buy N&C immediately.

• Mermin first if you are a pure CS student. Mermin's book is shorter, more focused, and better matched to a CS-undergraduate prerequisite set. Finish Mermin, then start N&C as the deeper reference.

• Watrous first is almost never right. The one exception is mathematics students with strong functional-analysis backgrounds who are specifically interested in information theory; they sometimes find Watrous more natural than the pedagogical textbooks.

The Indian reading environment

One practical feature of reading from India: your community of co-readers is thinner than at MIT or ETH Zürich, which raises the bar for self-discipline but also makes online communities important. QWorld India, Qiskit Fall Fest, the IBM Quantum Network at Indian partner institutions (IIT Madras, IIT Kanpur, IISc, TIFR), and the emerging National Quantum Mission fellowships all provide reading groups and discussion forums. If you are reading alone, try to find one other person — online or in your college — reading the same chapter at the same time. A weekly 30-minute call to discuss confusion points turns self-study from lonely into tractable.

The books you will not read but should know exist

Beyond the triad and the honorable mentions, a few specialist books are worth knowing about:

• Gottesman, Surviving as a Quantum Computer in a Classical World — a forthcoming QEC monograph. Drafts circulate online.
• Kitaev, Shen, Vyalyi, Classical and Quantum Computation (AMS, 2002). The best treatment of quantum complexity theory. Short, dense, mathematically rigorous.
• Wilde, Quantum Information Theory (Cambridge, 2nd ed 2017). A competitor to Watrous, broader in scope, also freely available on arXiv. Some people prefer Wilde; most people own both.
• Koenig and Smolin, forthcoming on quantum cryptography.

You will not read these for a long time, possibly ever. But knowing they exist lets you navigate the research literature — citations will send you to Wilde or Kitaev-Shen-Vyalyi and you should not be confused about what they are.

Where this leads next

• IBM Quantum Learning — the free video and hands-on platform that complements the books.
• Reading an arXiv Paper — the research-literature companion chapter.
• Qiskit Cirq PennyLane CUDA-Q — the SDK landscape for implementing what you read.
• The Landscape in 2026 — the ecosystem map, to see which parts of the field each book covers.
• What is Quantum Computing? — chapter 1 of this track, useful for recalibrating between chapters of the bigger textbooks.

References

1. Nielsen and Chuang, Quantum Computation and Quantum Information, 10th anniversary ed — Cambridge University Press.
2. John Preskill, Lecture Notes on Quantum Computation (Ph229) — theory.caltech.edu/~preskill/ph229. Free PDFs by chapter.
3. John Watrous, The Theory of Quantum Information (Cambridge, 2018) — free PDF at cs.uwaterloo.ca/~watrous/TQI.
4. N. David Mermin, Quantum Computer Science: An Introduction — Cambridge University Press.
5. NPTEL, Quantum Computing course listings — Indian faculty lectures, free, often with companion notes in English.
6. Wikipedia, Quantum Computation and Quantum Information — the book's own page, with a chapter-by-chapter outline and reception summary.