The Sycamore quantum processor was designed for quantum supremacy experiments and small-scale noisy intermediate-scale quantum (NISQ) applications. It supports quantum error correction based on the surface code and aims for 0.1% error two-qubit gates for error correction, while achieving quantum supremacy with 0.3-0.6% error rates. The architecture uses tunable transmon qubits with direct, tunable coupling to achieve short two-qubit gate times. The device uses adjustable couplers to manage coupling strength and minimize residual control errors. The Sycamore processor contains 142 transmon qubits, with 54 individually controlled and readout qubits and 88 adjustable couplers. The qubits are coupled through direct capacitive and indirect coupling mediated by couplers. The processor is fabricated on two high-resistivity silicon wafers and connected to a 3-layer Al-plated circuit board. The device uses custom-built multichannel digital-to-analog converters (DACs) and microwave arbitrary waveform generators (Microwave AWGs) for control and readout. The qubit state is measured via dispersive interaction with a far-detuned harmonic resonator. The readout signal is amplified and demodulated to infer the qubit state. The device uses cross entropy benchmarking (XEB) to estimate the fidelity of random quantum circuits with a large number of qubits. XEB is based on the observation that the measurement probabilities of a random quantum state have a similar pattern to laser "speckles". The XEB method uses numerical simulations to calculate the likelihood of a set of bitstrings obtained in an experiment according to the ideal expected probabilities. The theory behind XEB is described, including the use of linear and logarithmic versions of XEB to estimate the fidelity of random quantum circuits. The XEB method is used to estimate the fidelity of random quantum circuits with a large number of qubits, and the results are compared with other methods such as randomized benchmarking and Speckle purity benchmarking. The XEB method is also used to quantify errors in the quantum processor, including measurement errors and state preparation errors. The calibration process involves a series of experiments to learn optimal control parameters, including device configuration, single-qubit calibration, and two-qubit gate metrology. The calibration procedure is automated to systematize and optimize the calibration process, allowing for the comparison of different calibration strategies to determine optimal strategies for time, performance, and reliability. The Sycamore processor is designed to be highly programmable, allowing for dynamic changes in qubit frequency and qubit-qubit coupling. This tunability enables the implementation of various control strategies and accounts for non-uniformities in the processor's parameters. However, these extra degrees of freedom introduce additional sources of decoherence and control errors, requiring careful calibration to maintain performance. The calibration process is challenging due to the need for precise control-pulse shaping, individualThe Sycamore quantum processor was designed for quantum supremacy experiments and small-scale noisy intermediate-scale quantum (NISQ) applications. It supports quantum error correction based on the surface code and aims for 0.1% error two-qubit gates for error correction, while achieving quantum supremacy with 0.3-0.6% error rates. The architecture uses tunable transmon qubits with direct, tunable coupling to achieve short two-qubit gate times. The device uses adjustable couplers to manage coupling strength and minimize residual control errors. The Sycamore processor contains 142 transmon qubits, with 54 individually controlled and readout qubits and 88 adjustable couplers. The qubits are coupled through direct capacitive and indirect coupling mediated by couplers. The processor is fabricated on two high-resistivity silicon wafers and connected to a 3-layer Al-plated circuit board. The device uses custom-built multichannel digital-to-analog converters (DACs) and microwave arbitrary waveform generators (Microwave AWGs) for control and readout. The qubit state is measured via dispersive interaction with a far-detuned harmonic resonator. The readout signal is amplified and demodulated to infer the qubit state. The device uses cross entropy benchmarking (XEB) to estimate the fidelity of random quantum circuits with a large number of qubits. XEB is based on the observation that the measurement probabilities of a random quantum state have a similar pattern to laser "speckles". The XEB method uses numerical simulations to calculate the likelihood of a set of bitstrings obtained in an experiment according to the ideal expected probabilities. The theory behind XEB is described, including the use of linear and logarithmic versions of XEB to estimate the fidelity of random quantum circuits. The XEB method is used to estimate the fidelity of random quantum circuits with a large number of qubits, and the results are compared with other methods such as randomized benchmarking and Speckle purity benchmarking. The XEB method is also used to quantify errors in the quantum processor, including measurement errors and state preparation errors. The calibration process involves a series of experiments to learn optimal control parameters, including device configuration, single-qubit calibration, and two-qubit gate metrology. The calibration procedure is automated to systematize and optimize the calibration process, allowing for the comparison of different calibration strategies to determine optimal strategies for time, performance, and reliability. The Sycamore processor is designed to be highly programmable, allowing for dynamic changes in qubit frequency and qubit-qubit coupling. This tunability enables the implementation of various control strategies and accounts for non-uniformities in the processor's parameters. However, these extra degrees of freedom introduce additional sources of decoherence and control errors, requiring careful calibration to maintain performance. The calibration process is challenging due to the need for precise control-pulse shaping, individual