About Google
Google Quantum AI has produced some of the defining results in quantum computing, from the 2019 Sycamore experiment that ran the landmark quantum supremacy benchmark to the 2024 Willow chip, the first processor to show error-corrected qubits improving exponentially as the system scales.
Cirq is Google's open-source framework for programming quantum computers, built for today's noisy intermediate-scale (NISQ) devices where hardware details matter. It is the same framework Google uses to run experiments on its own quantum processors.
qsim is the ecosystem's optimized full state-vector simulator, written in C++ with gate fusion, AVX vectorized instructions, and OpenMP multithreading. Google used qsim to produce the cross-entropy benchmark results in its 2019 quantum supremacy paper. Memory doubles with each added qubit: Google's documentation puts a 30-qubit circuit at about 16 GB of RAM. For circuits too large for a full state vector, Open Quantum also offers a matrix product state engine built on quimb, which compresses the quantum state instead of storing every amplitude.