Architecture and design tradeoffs of Dwave Systems Superconducting quantum annealing processor

Dwave has developed a quantum annealing processor, based on an array of tunably coupled rf-SQUID flux qubits, fabricated in a superconducting integrated circuit process. Implementing this type of processor at a scale of 512 qubits [Dwave now has a 1024 qubit processor in its labs] and 1472 programmable inter-qubit couplers and operating at ~ 20 mK has required attention to a number of considerations that one may ignore at the smaller scale of a few dozen or so devices. Here Dwave discusses some of these considerations, and the delicate balance necessary for the construction of a practical processor that respects the demanding physical requirements imposed by a quantum algorithm. In particular Dwave reviews some of the design trade-offs at play in the floor-planning of the physical layout, driven by the desire to have an algorithmically useful set of inter-qubit couplers, and the simultaneous need to embed programmable control circuitry into the processor fabric. In this context Dwave has developed a new ultra-low power embedded superconducting digital-to-analog flux converters (DACs) used to program the processor with zero static power dissipation, optimized to achieve maximum flux storage density per unit area. The 512 single-stage, 3520 two-stage, and 512 three-stage flux-DACs are controlled with an XYZ addressing scheme requiring 56 wires. Our estimate of on-chip dissipated energy for worst-case reprogramming of the whole processor is ~ 65 fJ. Several chips based on this architecture have been fabricated and operated successfully at Dwave’s facility, as well as two outside facilities.

Arxiv – Architectural considerations in the design of a superconducting quantum annealing processor

Dwave describes how, starting with top-level requirements of a processor implementing a quantum annealing algorithm, we have designed its hardware graph and required control infrastructure, which allowed us to successfully operate processors with up to 512 rf-SQUID qubits using only 56 control lines for problem programming.

The most important feature of the new Dwave design is its zero static power dissipation – unlike traditional SFQ (single flux quantum) circuitry, which incorporates on-chip resistive current sources tapping a common voltage rail, this design biases all devices serially with a fixed current whose magnitude is set by a room-temperature resistor.

Complete reprogramming of all 9216 stages moving from -16 to +16 SFQ in their storage loops would dissipate on chip only about 65 femtojoules.

Chimera unit cell topology. (Left) Layout sketch: qubit bodies are represented by the elongated loops that span the whole width/height of the unit tile. Each qubit is coupled to four others within the unit tile via the internal coupler bodies (dark L-shaped objects). Qubits are coupled to others in neighboring tiles via external couplers (lighter dashed rectangles). Control circuitry (*-DACs and corresponding analog control structures) are placed within light-shaded areas between the qubit/coupler bodies. (Right) Graph representation: each unit tile corresponds to a complete bipartite graph K4,4(dark nodes and solid line edges). Qubits from different tiles are coupled in square grid fashion (dashed edges).

If you liked this article, please give it a quick review on ycombinator or StumbleUpon. Thanks