Affiliate disclosure

Book titles on this page link to Amazon. As an Amazon Associate, DataField.Dev earns from qualifying purchases — at no additional cost to you.

Further Reading: Google Cirq

Tagged Tier 1 (confident it exists and recommended) and Tier 2 (real and worth seeking, but verify the current version or URL).

The primary sources

  • The Cirq documentation, starting with the "Circuits" and "Simulation" guides. The moments model is explained better in the official docs than anywhere else, because it is Cirq's central design decision and the authors know it. Read the InsertStrategy page carefully — Case Study 2 exists because its per-call scoping is easy to skim past. Tier 1.
  • The Cirq API reference for cirq.Circuit, cirq.Moment, and cirq.InsertStrategy. Worth reading as reference rather than tutorial: the guarantees a Moment makes (disjoint qubits, same time slice) are the whole abstraction. Tier 1.
  • The cirq-google documentation on devices and GridDevice. What a Cirq device object contains — topology, gate sets, validation rules — and, by omission, what it does not: the calibration data Chapter 12 built its layout scoring on. Tier 1.
  • The Cirq "Comparison to other frameworks" material, if current. Vendor comparisons age badly and are written by interested parties, so read them alongside Chapter 18 rather than instead of it. Tier 2.

On the design decisions

  • Any careful treatment of quantum circuit scheduling and ASAP/ALAP. Cirq's moments and Qiskit's scheduling passes are two answers to the same question, and reading about the general problem makes both look like choices rather than conventions. Connects directly to Chapter 13 §13.6, where ALAP scheduling silently moved a circuit's idle time. Tier 2.
  • The OpenQASM 3 specification's treatment of timing and delay (Chapter 6, Appendix F). A third position on the same question: an interchange format that makes timing expressible but optional. Tier 1.
  • The sympy documentation on symbols and expressions. Cirq's parameters are just sympy, so the full CAS is available inside a gate argument. Ten minutes here buys more than any Cirq-specific parameter tutorial. Tier 1.

On endianness and conventions

  • Nielsen and Chuang's notational conventions, and any two other quantum computing texts. Compare how each writes a multi-qubit basis state. The disagreement is not a Cirq-versus-Qiskit quirk; it is a field-wide convention split that the frameworks inherited. Tier 1.
  • Chapter 3 §3.5 and Chapter 4 of this book, on little-endian ordering. Worth rereading now that you have seen the alternative — several passages that read as statements about quantum computing turn out to be statements about Qiskit. Tier 1.
  • Any discussion of byte-order bugs in classical systems programming. The failure mode in Case Study 1 — a convention mismatch at a boundary, invisible to symmetric test data, its own inverse so that double-conversion hides it — is exactly the classical bug, and the classical literature on it is fifty years deeper. Tier 2.

On what Cirq was built for

  • Arute et al., "Quantum supremacy using a programmable superconducting processor" (2019), Nature 574, 505. The Sycamore result, run through Cirq. Read it for what the framework was designed around: a specific 2-D grid device, precise timing control, and random circuit sampling. The design choices in this chapter stop looking arbitrary. Pair it with Chapter 1 §1.5 on the classical counterattack. Tier 1.
  • OpenFermion — the chemistry library built on Cirq. If your interest is Chapter 36's territory, this is the strongest reason to use Cirq specifically. Tier 1.
  • Qsim, Google's high-performance simulator. Substantially faster than the default Cirq simulator for large circuits, and the natural next step after Chapter 11's simulation-method comparison. Tier 1.

Forward references

  • Chapter 15 — Q#, which is a language rather than a library, and takes the strictness idea considerably further than Cirq does.
  • Chapter 16 — PennyLane, whose parameter handling makes Cirq's sympy symbols look conservative, and which differentiates through circuits.
  • Chapter 18 — the full comparison, including the translation table this chapter only sketched.
  • Appendix E — the gate-by-gate, call-by-call translation dictionary across all five frameworks.

Where to go next. If one thing: the Cirq docs' Circuits guide, read specifically for the moments model. It is a genuinely different way to think about a circuit, and it will change how you read Qiskit code even if you never write Cirq again.

Then Chapter 15 — Q#, where the compiler stops being a translator and starts being an editor with opinions.