A quantum computing testbed based on trapped ions that is available to the research community as an open platform for a range of quantum computing applications.
Quantum information processing has reached an inflection point transitioning from proof of principle scientific experiments to small noisy quantum processors. To accelerate this process, it is necessary to provide the scientific community with access to testbed systems that provide full specifications, enable low-level access to native gate implementations, make vertical integration approaches possible, and provide ways to fully specify the scheduling of gates. Access to noisy intermediate-scale quantum (NISQ) systems is needed to understand and optimize the noise properties, learn how to characterize and validate quantum operation, and to incubate the development and optimization of quantum algorithms for scientific applications.
The Quantum Scientific Computing Open User Testbed (QSCOUT) is a 5-year DOE program funded by the Office of Science’s Advanced Scientific Computing Research (ASCR) program to build a quantum testbed based on trapped ions that is available to the research community. As an open platform, it will not only provide full specifications and control for the realization of all high- level quantum and classical processes, it will also enable researchers to investigate, alter, and optimize the internals of the testbed and test more advanced implementations of quantum operations. QSCOUT will be made operational in stages, with each stage adding more ion qubits, greater classical control, and improved fidelities. We will leverage the specific strengths of trapped ion systems: the identical qubits with long qubit coherence times, the high-fidelity single and multi-qubit operations possible in these systems, the low cross-talk addressing of individual qubits in the register, and the all-to-all connectivity available in trapped ion quantum registers. In the first stage, we will make a quantum register of 3 qubits available. Parallel single qubit gates and sequential two-qubit Mølmer-Sørensen gates between any pair of qubits will be available. Target fidelities for single qubit operations are 99.5%, target fidelities for two-qubit gates are 98%. At the beginning of a computation, each quantum bit is prepared in the |0〉 state of the z-basis. At the end of a computation the entire quantum register is measured in the z-basis. For each measurement of the quantum register, the state of each qubit will be available to users.
QSCOUT is led by Dr. Susan Clark at Sandia in collaboration with a team of AMO physicists, quantum computing theorists, engineers, and computer scientists. In addition, Peter Love (Tufts University) and Ken Brown (Duke University) provide theory support.
The QSCOUT hardware will be realized as a trapped ion system. A chain of ytterbium ions will be stored in a Sandia surface ion trap, which offers excellent optical access for state preparation, detection and qubit manipulations. Qubits are encoded in the hyperfine clock states of each ytterbium-171 ion and a chain of ions serves as the qubit register. Single- and multi-qubit operations are implemented with optical Raman transitions using a 355nm pulsed laser. Imaging of an acousto-optical modulator (AOM) array onto the ion chain is used to realize individual addressing of qubits in the register. At the end of a computation, the quantum state of each qubit in the register will be read out and reported for each qubit and each detection event. This is achieved with standard fluorescence detection by imaging the chain of ions on an array of multi-mode fibers connected to an array of individual photomultiplier tubes.
The first available system, QSCOUT Testbed 1.0
In the first stage, we will make a quantum register of 3 qubits available. Parallel single qubit gates and sequential two-qubit Mølmer-Sørensen gates between any pair of qubits will be available. Target fidelities for single qubit operations are 99.5%, target fidelities for two-qubit gates are 98%. At the beginning of a computation, each quantum bit is prepared in the |0〉 state of the z-basis. At the end of a computation the entire quantum register is measured in the z-basis. For each measurement of the quantum register, the state of each qubit will be available to users.
The system will be characterized using validation and verification methods developed by Sandia's Quantum Performance team. This will include characterizations of single and two-qubit gates using Gate Set Tomography (GST). Special attention will be made to reduce non-Markovian errors such as drifts and context-dependent gate errors. The results of these characterizations will be made available to users.
In addition to a standard Gate Level access, QSCOUT will enable users to propose and use alternate pulse shapes to realize gates or enable them to realize additional native gates by specifying the pulse sequence needed for their implementation.
Pulse sequences are realized using a custom multi-tone Direct Digital Synthesizer (DDS) system where the frequency, phase and amplitude modulations of all tones can be specified as spline functions.
Pulse sequences can be streamed out in real time and enable the user to run an unlimited number of gate sequences efficiently. Users are encouraged to develop gate implementations as pulse shapes in close collaboration with Sandia scientists who will implement the pulse shapes on the testbed system.
Jaqal™ the Quantum Assembly Language for QSCOUT
Just Another Quantum Assembly Language (Jaqal) is the programming language used to specify programs executed on QSCOUT. This document contains a specification of Jaqal along with a summary of QSCOUT 1.0 capabilities, for example, Jaqal programs, and plans for possible future extensions.