Google's latest results in quantum AI! Exponentially accelerated quantum algorithms for classical mechanics

While the IBM Quantum Summit and QC Ware's Q2B Silicon Valley conference dominated last week's news stream, other quantum news has been popping up. Below, we briefly review other highlights of coverage of developments in the quantum field.

First up is research published last Thursday in the journal Science, in which two groups of researchers at Harvard and Princeton University utilized entire molecules as quantum bits. The two groups of researchers used optical tweezers to control calcium fluoride molecules to achieve entanglement. A review in Nature by Davide Castelvecchi summarizes the two studies well:

"Both studies used arrays of optical tweezers, with one molecule captured in each optical tweezer unit. Using laser technology, they cooled the molecules to a temperature of tens of microkelvins, just one millionth of absolute zero. In this state, the molecules are close to being completely still. They can either stop rotating or rotate with only one quantum of angular momentum (called ħ), which is the smallest rotational frequency they can possibly have."

"Both teams used non-rotating molecules to represent the |0⟩ state of a quantum bit and rotating molecules to represent the |1⟩ state."

So there may be one more member in the queue of quantum bits.

Years ago, there were attempts to use full molecules as quantum bits; these two recent efforts are a big step forward. There are, of course, a growing number of quantum bits that are vying for the market: notably superconductivity, trapped ions, neutral atoms, photonics, silicon spins, diamond vacancies, and topology.

For most applications, molecular quantum computers will be slower than computers using other types of quantum bits, the researchers say. But molecules may be a natural environment for manipulating quantum information using "quantum bits," which have three possible states: |-1⟩, |0⟩ and |+1⟩. Such Qutrits could provide a way to perform quantum simulations of complex materials or fundamental forces in physics.

In fact, back in August of last year, Peter Chapman, an expert in captured ions and CEO of IonQ, predicted a whole new paradigm breakthrough for quantum bits.

"It's a bit controversial, but I think there are other quantum bit modes that could be better in the long run. It's just that in the next five to ten years, they won't be at the level that we [capture ion quantum bits] are at now. But if you said to me, which quantum bit mode might I be using in 15 or 20 years? I don't know, neutral atoms might be an interesting platform at that point, or something."

New product hints at market prospects for (small) commercial QPUs

Also on Thursday, Rigetti Computing introduced the 9-qubit Novera QPU for sale directly to researchers. 

Rigetti reports, "The Novera QPU can be ordered at rigetti.com/novera starting at $900,000 and will ship within 4-6 weeks after order confirmation, shipping and logistics are finalized."

It's an interesting bet and could be a prudent attempt at the commercial QPU market. Here's a brief summary of Rigetti's use of the new QPU:

"The Novera QPU enables general-purpose, gate-based quantum computation that can be used by quantum software and algorithm experts to prototype and test (1) hybrid quantum algorithms, (2) characterization, calibration, and error mitigation, and (3) quantum error correction (QEC) experiments."

"In addition, organizations wishing to develop components of their quantum computing stacks can leverage the Novera QPU to accelerate work in the areas of (1) control electronics and software, (2) QEC decoders, (3) control optimization algorithms, (3) native gate architectures, and (4) measurements and calibrations and accompanying software."

This sounds more like a research tool than a commercial component for self-built quantum computing, and Rigetti reports that the Novera QPUs are manufactured at "Rigetti's Fab-1 facility - the industry's first dedicated integrated quantum device manufacturing facility ".

Here is Rigetti's description of the Novera QPU components:

- A puck containing a 9-quantum-bit and 5-quantum-bit chip, interposers and printed circuit boards for transmitting signals to SMPM connectors on the periphery of the puck.

- A tower suspended from the MXC to connect the coaxial cable between the Puck and the SMA patch panel; the tower transfers cooling power from the MXC to the chip.

- A shield that surrounds the tower to isolate the chip from infrared radiation and stray magnetic fields.

- Payload brackets and a signal chain mounted around the tower, including signal conditioning equipment such as ferrite isolators, diplexers, filters and optional quantum limiting amplifiers.

Novera utilizes the same architecture as Rigetti's fourth-generation Ankaa-class architecture, featuring tunable couplers and a square lattice for denser connectivity and fast 2-qubit operation.

Notably, a number of related QPUs have been commissioned by the National Labs.

Google's classical mechanics quantum algorithm achieves an exponential speedup

There is currently a feverish effort to develop quantum algorithms - like Peter Shaw's algorithm - that can prove faster than classical algorithms.

Google reported such an advance in a blog post last week with a paper ("Exponential Quantum Speedup in Simulating Coupled Classical Oscillators").

A Google team reports the discovery of a new quantum algorithm that provides an exponential advantage for modeling coupled classical harmonic oscillators (harmonic oscillators). These are some of the most fundamental and pervasive systems in nature and can characterize the physics of countless natural systems from electrical circuits and molecular vibrations to bridge mechanics.

Google, in collaboration with scientists at Macquarie University and the University of Toronto, has discovered a "mapping" that can transform any system involving coupled oscillators into a problem that describes the time evolution of quantum systems. Under certain constraints, a quantum computer can solve this problem several times faster than a classical computer.

In addition, the collaborating team used this mapping to show that any problem that can be solved efficiently by a quantum algorithm can be recast as a problem involving a network of coupled oscillators, albeit in exponential numbers. In addition to unveiling previously unknown application areas for quantum computers, this result provides a new way to design new quantum algorithms by purely reasoning about classical systems.

48 Logic Quantum Bits Perform Error Correction Algorithm

The researchers cite the following highlights:

- Creation and entanglement of the largest logic quantum bits to date, showing a code distance of 7, which enables detection and correction of arbitrary errors that occur during the operation of entangled logic gates (the larger the code distance, the greater the resistance to quantum errors).

Furthermore, it was shown for the first time that increasing the code distance can indeed reduce the error rate in logic operations.

- Forty-eight small logic quantum bits for executing complex algorithms were implemented, and their performance exceeded that of the same algorithms when executed using physical quantum bits.

- By controlling 280 physical bits, 40 medium-sized error-correcting codes were constructed.

Nature reported on this work (Logic quantum processor based on reconfigurable atomic arrays) last week. Here is the abstract:

"We report the implementation of a programmable quantum processor based on encoded logic quantum bits running up to 280 physical quantum bits. Utilizing logic-level control and a partitioned architecture in a reconfigurable array of neutral atoms, our system combines high two-qubit gate fidelity, arbitrary connectivity, and fully programmable single-qubit rotation and mid-circuit readout. By running this logic processor with various types of codes, we demonstrate improved dual-qubit logic gates by scaling the surface code distance from d = 3 to d = 7, preparation of color-coded quantum bits with break-even fidelity, fault-tolerant creation of logic GHZ states and feedforward entangled stealthy transitions of states, and operation with 40 color-coded quantum bits."

"Finally, using three-dimensional code blocks, we implemented complex computational sampling circuits with up to 48 logic quantum bits that are connected and entangled via 228 logic double quantum bit gates and 48 logic CCZ gates. We found that this logic coding greatly improves the algorithmic performance of error detection, outperforming physical quantum bit fidelity in both cross-entropy benchmark tests and quantum simulations of fast scrambling codes."

The researchers say these results herald "the arrival of early error-correcting quantum computing and point the way to large-scale logic processors."

Optimized control and enhanced laser power should allow such architectures to reach 10,000 physical bits, so there should be considerable headroom, the paper notes; and, since all control operations are done via the laser, it should be possible to use photonic links to connect independent hardware.

With this approach, scientists would not need thousands, hundreds of thousands, or millions of physical quantum bits to correct errors. If this method works, the speed of error correction will be amazingly fast.