D-Wave Systems, the world’s first quantum computing company, announced that it has broken the 1000 qubit barrier, developing a processor about double the size of D-Wave’s previous generation and far exceeding the number of qubits ever developed by D-Wave or any other quantum effort. This is a major technological and scientific achievement that will allow significantly more complex computational problems to be solved than was possible on any previous quantum computer.
D-Wave’s quantum computer runs a quantum annealing algorithm to find the lowest points, corresponding to optimal or near optimal solutions, in a virtual “energy landscape.” Every additional qubit doubles the search space of the processor. At 1000 qubits, the new processor considers 2^1000 possibilities simultaneously, a search space which dwarfs the 2^512 possibilities available to the 512-qubit D-Wave Two. In fact, the new search space contains far more possibilities than there are particles in the observable universe.
As the only manufacturer of scalable quantum processors, D-Wave breaks new ground with every succeeding generation it develops. The new processors, comprising over 128,000 Josephson tunnel junctions, are believed to be the most complex superconductor integrated circuits ever successfully yielded. They are fabricated in part at D-Wave’s facilities in Palo Alto, CA and at Cypress Semiconductor’s wafer foundry located in Bloomington, Minnesota.
“Temperature, noise, and precision all play a profound role in how well quantum processors solve problems. Beyond scaling up the technology by doubling the number of qubits, we also achieved key technology advances prioritized around their impact on performance,” said Jeremy Hilton, D-Wave vice president, processor development. “We expect to release benchmarking data that demonstrate new levels of performance later this year.”
As far as the exact number of qubits in the new processor, it varies somewhat for each specific processor. The 1,000+ qubit processor was designed to deliver a 1,152 qubit region out of a complete 2,048 qubit fabric. The additional qubits give us flexibility to select the ‘sweet spot’ of the processor, which results from the analog nature of the quantum devices and their sensitivity to parametric variation across the processor. As with previous generations, magnetic offsets and manufacturing variability inherent in the superconductor circuit fabrication push some of these qubits out of the performance range. Those qubits are disabled during the qualification process. In all cases, these processors will have significantly more than 1,000 qubits.
The 1000-qubit milestone is the result of intensive research and development by D-Wave and reflects a triumph over a variety of design challenges aimed at enhancing performance and boosting solution quality. Beyond the much larger number of qubits, other significant innovations include:
Lower Operating Temperature
While the previous generation processor ran at a temperature close to absolute zero, the new processor runs 40% colder. The lower operating temperature enhances the importance of quantum effects, which increases the ability to discriminate the best result from a collection of good candidates.
Through a combination of improved design, architectural enhancements and materials changes, noise levels have been reduced by 50% in comparison to the previous generation. The lower noise environment enhances problem-solving performance while boosting reliability and stability.
Increased Control Circuitry Precision
In the testing to date, the increased precision coupled with the noise reduction has demonstrated improved precision by up to 40%. To accomplish both while also improving manufacturing yield is a significant achievement.
The new processors comprise over 128,000 Josephson junctions (tunnel junctions with superconducting electrodes) in a 6-metal layer planar process with 0.25μm features, believed to be the most complex superconductor integrated circuits ever built.
New Modes of Use
The new technology expands the boundaries of ways to exploit quantum resources. In addition to performing discrete optimization like its predecessor, firmware and software upgrades will make it easier to use the system for sampling applications.
SOURCES – DWave systems