keepcalculator

06 Numbers and Powers

Google Quantum Supremacy: Fact Check

Card No. 06·73316 · Updated · Standard formula, shown below

Did Google achieve quantum supremacy? Yes, in 2019 Google announced that its Sycamore processor performed a specific computational task in 200 seconds that would take a state-of-the-art supercomputer approximately 10,000 years, marking the first claimed demonstration of quantum supremacy. However, the claim has been challenged and refined, with subsequent classical algorithms and hardware improvements shrinking the gap, so the current consensus is that Google demonstrated a significant milestone, but the term “supremacy” remains debated.

What Is Quantum Supremacy?

Quantum supremacy is the hypothetical point at which a quantum computer can perform a calculation that is practically impossible for any classical computer, regardless of how powerful. The term was coined by John Preskill in 2012. It does not mean the quantum computer is useful for everyday tasks; rather, it is a scientific benchmark proving that quantum hardware can outperform classical hardware on at least one specific problem.

The problem chosen by Google was to sample the output of a pseudo-random quantum circuit. This is a task that has no practical application but is designed to be extremely hard for classical computers to simulate. The circuit generates a probability distribution that is exponentially complex to compute classically, making it a perfect candidate for demonstrating quantum advantage.

The 2019 Sycamore Experiment

On October 23, 2019, a team led by John Martinis at Google AI Quantum published a paper in Nature titled “Quantum supremacy using a programmable superconducting processor.” They used a 53-qubit processor named Sycamore, which is based on superconducting transmon qubits arranged in a two-dimensional grid.

The experiment involved running a random circuit with 53 qubits and 20 cycles of gates. The output was a sequence of 53 bits, and the task was to sample from the output distribution. The team estimated that the same task would take the world’s fastest supercomputer, Summit at Oak Ridge National Laboratory, about 10,000 years. Google’s Sycamore completed the task in about 200 seconds.

How It Works: The Random Circuit Sampling Problem

The core of the experiment is a computational problem called random circuit sampling (RCS). Here’s a simplified breakdown:

  • Qubits: Sycamore has 53 qubits, which can exist in superposition of 0 and 1 simultaneously.
  • Gates: The qubits are manipulated with a sequence of quantum gates, including single-qubit rotations and two-qubit controlled-NOT gates.
  • Random Circuit: The gates are arranged in a random but known pattern, creating a highly entangled state.
  • Measurement: Each qubit is measured, collapsing the superposition to a classical bit string of 0s and 1s.
  • Sampling: The process is repeated many times to obtain a set of bit strings that follow a specific probability distribution.

Classically, simulating this distribution requires computing the probability of each possible bit string, which scales exponentially with the number of qubits. For 53 qubits, there are 2^53 (about 9 quadrillion) possible outcomes, making exact simulation infeasible.

Challenges and Counterclaims

Immediately after Google’s announcement, IBM challenged the claim. IBM argued that the 10,000-year estimate was overly conservative and that a classical supercomputer could simulate the same task in about 2.5 days using a technique called “schmidt decomposition.” While 2.5 days is still much slower than 200 seconds, it undermines the notion of “supremacy” as a hard wall.

In 2021, a team from the Chinese Academy of Sciences led by Jian-Wei Pan used a photonic quantum computer named Jiuzhang to perform Gaussian boson sampling, another classically hard problem. They claimed a speedup of 10^14 times over classical simulation, which is even more dramatic than Google’s claim.

More recently, in 2023, researchers at IBM and other institutions developed improved classical algorithms that can simulate random circuit sampling more efficiently. For example, a tensor network method can simulate a 53-qubit random circuit in about 15 hours on a large cluster. This reduces the quantum advantage but does not eliminate it.

What Does This Mean for Quantum Computing?

Quantum supremacy is a scientific milestone, not a practical one. The random circuit sampling problem has no real-world use. The achievement demonstrates that quantum computers can perform certain tasks faster than classical ones, but it does not mean quantum computers are ready for commercial applications.

Practical quantum advantage will require fault-tolerant quantum computers with thousands of logical qubits, which are still years away. Current quantum processors are noisy and error-prone, limiting their usefulness. However, the 2019 experiment provided valuable data on error rates and control, guiding future development.

It is also important to note that quantum supremacy is not a single event but a moving target. As classical algorithms improve, the bar for quantum advantage rises. The term “quantum advantage” is often preferred because it is less absolute and acknowledges that the gap may narrow over time.

Common Misconceptions

Several misconceptions surround the Google claim:

  • Misconception: Quantum computers can now solve any problem faster. Truth: Sycamore was designed for one specific task. It cannot run Shor’s algorithm or break encryption.
  • Misconception: The 10,000-year estimate is exact. Truth: It was an extrapolation based on the best classical algorithm at the time. Subsequent algorithms have reduced it.
  • Misconception: Quantum supremacy means quantum computers are useful. Truth: It is a proof-of-principle, not a practical tool.
  • Misconception: IBM’s rebuttal disproved the claim. Truth: It showed the gap was smaller than claimed, but still a significant speedup existed.

Frequently Asked Questions

Did Google actually achieve quantum supremacy?

Yes, Google’s 2019 experiment is widely recognized as the first demonstration of quantum supremacy, though the term is debated. The experiment performed a specific task faster than any classical computer could at the time. However, subsequent classical algorithms have narrowed the gap, so it is not an absolute, permanent achievement.

What is the difference between quantum supremacy and quantum advantage?

Quantum supremacy is the older term, implying a clear, unambiguous victory over classical computers. Quantum advantage is a more nuanced term that acknowledges the speedup may be for a specific task and may not be overwhelming. Many researchers now prefer “quantum advantage” because it is more precise.

Can quantum computers break encryption now?

No. Breaking RSA or ECC encryption requires thousands of logical qubits with error correction, which does not exist yet. Current quantum computers have only tens to hundreds of noisy physical qubits. The 2019 experiment did not involve any cryptographic algorithms.

How long would it take a classical computer to do what Sycamore did?

Google estimated 10,000 years on Summit. IBM later argued 2.5 days with a different method. In 2023, improved tensor network methods reduced it to about 15 hours on a large cluster. So it varies widely depending on the algorithm and hardware.

What is the next milestone after quantum supremacy?

The next major goal is fault-tolerant quantum computing, where errors are corrected and logical qubits can run longer algorithms. Companies like IBM and Google have roadmaps to achieve this by the late 2020s or early 2030s. Another milestone is demonstrating a practical quantum advantage in a useful application like chemistry simulation.

Is quantum supremacy a hoax?

No. The experiment was peer-reviewed and published in Nature. The results have been replicated and extended by other groups. The debate is about the magnitude of the speedup and the definition of supremacy, not about whether the experiment was real.

Related

Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.

More cards in this drawer

The whole Numbers and Powers drawer →