How reality gets in the way of quantum computing hype

SEP 01, 2022

Baidu is the latest entrant in the quantum computing race, which has been ongoing for years among both big tech and startups. Nevertheless, quantum computing may face a trough of disillusionment as practical applications remain far from reality.

Baidu makes its quantum move

Last week, Baidu unveiled its first quantum computer, coined Qian Shi, as well as what it claimed is the world’s first “all-platform integration solution,” called Liang Xi. The quantum computer is based on superconducting qubits, which is one of the first types of qubits, among many techniques that have been investigated, that became widely adopted, most notably in the quantum computer which Google used to proclaim quantum supremacy.

How to re-engineer global platforms for our multi-cloud reality_

Qian Shi has a computing power of 10 high-fidelity qubits. High fidelity refers to low error rates. According to the Department of Energy’s Office of Science, once the error rate is less than a certain threshold — i.e., about 1% — quantum error correction can, in theory, reduce it even further. Beating this threshold is a milestone for any qubit technology, according to the DOE’s report.

Further, Baidu said it has also completed the design of a 36-qubit chip with couplers, which offers a way to reduce errors. Baidu said its quantum computer integrates both hardware, software and applications. The software-hardware integration allows access to quantum chips via mobile, PC and the cloud.

Moreover, Liang Xi, Baidu claims, can be plugged into both its own and third-party quantum computers. This may include quantum chips built on other technologies, with Baidu giving a trapped ion device developed by the Chinese Academy of Sciences as an example.

“With Qian Shi and Liang Xi, users can create quantum algorithms and use quantum computing power without developing their own quantum hardware, control systems or programming languages,” said Runyao Duan, director of the Institute for Quantum Computing at Baidu Research. “Baidu’s innovations make it possible to access quantum computing anytime and anywhere, even via smartphone. Baidu’s platform is also instantly compatible with a wide range of quantum chips.”

Despite Baidu’s claim of being the world’s first such solution, the Liang Xi platform is reminiscent of Israel’s Innovation Authority approach, which is also aimed at being compatible with various types of qubits.

Although this is Baidu’s first quantum computer, the company has already submitted over 200 patents throughout the last four years since the founding of its quantum computing research institute. The patents span various areas of research including quantum algorithms and applications, communications and networks, encryption and security, error correction, architecture, measurement and control and chip design.

Baidu claims its offering paves the way for the industrialization of quantum computing, making it the latest company to make grandiose claims about quantum computing being on the verge of widespread adoption. Some quantum startups have already amassed staggering valuations of over $1 billion.

However, real applications for quantum computers, besides encryption, have yet to emerge. And even if they do, it’s expected that those will require thousands, which is far from what has anyone yet been able to achieve. For example, this scalability concern led Intel to stop pursuing the popular superconducting qubit approach in favor of the less mature silicon and silicon-germanium qubits, which are based on transistor-like structures that can be manufactured using traditional semiconductor equipment.

Nevertheless, voices are already emerging to warn of overhyping the technology. In the words of the Gartner Hype Cycle, this may mean that quantum computing may approach its trough of disillusionment.

The other main challenge in quantum computing is that real qubits tend to be too noisy, leading to decoherence This leads to the necessity of using quantum error correction, which increases the number of qubits far above the theoretical minimum for a given application. A solution called noisy intermediate scale quantum (NISQ) has been proposed as a sort of midway, but its success has yet to be shown.

The history of classical computers is filled with examples of applications that the technology enabled that had never been thought of beforehand. This makes it tempting to think that quantum computing may similarly revolutionize civilization. However, most approaches for qubits currently rely on near-absolute zero temperature. This inherent barrier implies quantum computing may remain limited to enterprises.