Google Reignites IBM ‘Quantum Supremacy’ Feud with Breakthrough in Error-Free Computing
Google's recent quantum breakthrough may not be welcome news to IBM. | Credit: NurPhoto / Getty Images.
Key Takeaways
- Google and IBM are two of the leading players in quantum research.
- The two companies have taken different approaches to quantum error correction.
- Researchers recently demonstrated important progress in memory error correction in a breakthrough for Google’s surface code-based approach.
In 2019, Google made headlines with a paper entitled “Quantum Supremacy Using a Programmable Superconducting Processor.” The paper described a breakthrough in which researchers performed a target computation they claimed would take the world’s fastest supercomputer 10,000 years in just 200 seconds.
Other scientists have since disproved the claim by demonstrating equivalent and even faster performance with classical computers. However, several important advancements have occurred in the last five years, including a recent breakthrough in error correction by the same Google researchers.
What is Quantum Supremacy?
Quantum supremacy refers to the achievement of using quantum hardware to solve a problem that no classical computer could solve in a realistic timeframe.
Discussions of the technology often pit leading players in the space against each other, depicting the field as a race between companies like Google and IBM to achieve supremacy.
For example, after Google published the 2019 paper, an IBM research blog declared that although the experiment was “an excellent demonstration” of progress in quantum computing, “it should not be viewed as proof that quantum computers are ‘supreme’ over classical computers.”
Quantum Error Correction
In the quest for quantum supremacy, Google and IBM have both invested significant effort in developing new error correction techniques to mitigate the effects of quantum noise, i.e., unwanted disturbances that can affect quantum systems and cause errors in computations.
Reducing the error rate of quantum computers is seen as a necessary precondition for functional “fault-tolerant” machines.
Research teams at Google and IBM have taken two distinct approaches to quantum error correction: the former focused on surface codes, and the latter on cat codes.
Surface Codes vs. Cat Codes
Surface and cat codes represent two approaches to encoding multiple noisy physical qubits (quantum bits) into a single error-corrected logical qubit.
Surface codes use a two-dimensional grid of qubits that acts as a surface for error correction. They are known for their high threshold for error rates. However, scaling them for applied computation has proven difficult.
In contrast, cat codes leverage specific quantum states known as cat states to encode a logical qubit into a superposition of multiple physical qubits’ states. While they can be more resource-efficient than surface codes, they tend to have lower error thresholds.
A Breakthrough for Google
In theory, quantum error correction techniques should be able to reduce error rates exponentially by adding more physical qubits to logical qubits. However, this exponential error suppression only occurs if the physical error rate is below a critical threshold.
In a recent paper , Google researchers demonstrated two surface code memories operating below this threshold. They showed that a logical qubit composed of 105 physical qubits suppressed errors more effectively than a logical qubit composed of 72.
Of course, memory alone is only one aspect of computation. The next step is to perform logical operations on the information stored.
Nevertheless, the paper’s authors stressed the significance of their results, which they said “show that superconducting processors can remain stable over the hours-long timescales required for large-scale fault-tolerant algorithms.”
Alongside IBM’s parallel advancements , the latest breakthrough suggests that practical error correction could be within reach.
IBM Cuts China Research Team as Washington-Beijing Tensions Continue, 1,000 Jobs Affected
