What is Quantum Supremacy and When Was It Achieved?

The term "quantum supremacy" was introduced by theoretical physicist John Preskill in 2012. Scientifically, it represents a fundamental computational milestone. It occurs when a quantum device solves a specific problem in a reasonable time, while a classical supercomputer can theoretically solve it, but inefficiently, taking years or even thousands of years to produce a result.

Initial demonstrations of quantum supremacy were purely laboratory-based due to the susceptibility of systems to computational noise (errors). To prove the viability of the technology under such conditions, engineers had to use synthetic algorithms, such as random quantum circuit sampling (RCS).

These tests had no direct commercial value but served an important purpose: they established the superiority of quantum architecture over classical systems in a specific niche and paved the way for the industry to seek practical applications for the technology.

In 2019, a research team from Google first claimed to have achieved quantum supremacy. The Google Sycamore processor, featuring 53 superconducting qubits, completed an RCS task in 200 seconds. Researchers asserted that the most powerful classical supercomputer at the time, Summit, would take about 10,000 years to accomplish the same task.

IBM disputed Google's announcement, claiming that Summit could solve the problem in just two and a half days. According to IBM, if not only processors but also the vast amounts of RAM and disk space of the supercomputer were effectively utilized, the exponential complexity could be circumvented.

Later, Chinese research teams reported surpassing the milestone. They demonstrated supremacy on two different physical architectures: on the optical quantum computer Jiuzhang, which uses photons for boson sampling, and on upgraded superconducting systems with the QPU Zuchongzhi 3.0. In March 2025, the system generated a million samples in just a few minutes. According to the Chinese team, it would take the world's most powerful classical supercomputer, Frontier, about 6.4 billion years to accurately simulate this specific process.

Although tasks like RCS do not provide practical or commercial benefits, they play a crucial role: they demonstrate that as the number of quality qubits increases, quantum power becomes insurmountable for classical von Neumann architecture.

What is Quantum Utility?

Quantum utility (quantum utility) is achieved when quantum computers transition from being laboratory record generators to tools for scientific research. At this stage of development, quantum systems do not outperform supercomputers across all metrics but can explore physical problems at a scale that is inaccessible to direct classical modeling.

Quantum utility represents the maximum capability of quantum computers in the NISQ era. To move to the next stage (FTQC), engineers focus not on increasing the number of qubits but on error mitigation. This method allows for accurate calculations to be extracted from "noisy" systems before they lose their quantum state.

Error mitigation should be strictly distinguished from full hardware error correction, which signifies the next historical stage.

The concept was proposed and validated by IBM in 2023, effectively marking the beginning of the quantum utility period, which continued into 2026. In the experiment, a 127-qubit Eagle processor was used to model the properties of complex magnetic materials. Leveraging noise suppression methods, the processor produced results that could not be accurately calculated using classical methods.

To achieve quantum utility, a hybrid architecture is often employed, simultaneously utilizing QPU, CPU, and GPU. This balance allows for efficient distribution of computational tasks.

In May 2026, IBM, in collaboration with Cleveland Clinic and Japan's RIKEN Institute, modeled a massive protein-ligand complex consisting of 12,635 atoms using such heterogeneous computation. The task was solved on two quantum computers and two classical supercomputers.

What is Quantum Advantage?

The terms "quantum supremacy" and "quantum advantage" (quantum advantage) are often used interchangeably in the media, but in scientific and business contexts, they denote different historical stages of technology development.

While supremacy is a laboratory demonstration of the fundamental computational power of quantum hardware, advantage encompasses a set of conditions. It is achieved when a device solves a specific practical problem faster, cheaper, or more accurately than the best classical supercomputer.

The main criterion for advantage is practical and economic feasibility. Businesses do not need a complex and expensive QPU if a traditional cluster can model the behavior of a molecule for a new drug or calculate the properties of a super-strong alloy in a similar time and budget.

Achieving quantum advantage, along with FTQC, is a primary goal for leading technology companies and startups over the next three to four years.

Examples from roadmaps include:

  • IBM. By the end of 2026, the company plans to demonstrate "the first examples of practical quantum advantage" using the Nighthawk processor. It will be capable of executing deep circuits of 7,500 gates in close hybrid integration with classical supercomputers. By 2029, developers aim to release a full-scale FTQC system operating with 200 logical qubits — Starling;
  • QuEra Computing. This startup, specializing in neutral atom architecture, plans to release a system with 100 fault-tolerant logical qubits by 2026. Engineers estimate this capacity will be sufficient to begin solving the first commercially significant tasks in chemistry and materials science that are inaccessible to classical computers;
  • Quantinuum in collaboration with Microsoft. The company aims to achieve business goals by 2030. The focus is on releasing a fifth-generation Apollo quantum computer. The ion trap-based system is expected to achieve hundreds of logical qubits with deep error correction, integrating with AI platforms and Microsoft Azure Quantum cloud infrastructure;
  • Google Quantum AI. Following the presentation of the 105-qubit Willow processor at the end of 2024, the company achieved success in error suppression. The goal is to complete the development of a large-scale quantum computer with hardware noise correction capable of reliably processing data for commercial tasks by the end of this decade.
IBM's roadmap. Source: IBM.

In Which Areas Are Quantum Computing Most Effective?

Initial real results are being achieved exclusively in disciplines requiring the simulation of complex quantum-mechanical systems. Classical processors are inefficient at calculating molecular interactions: adding each new electron to a model causes exponential data growth. In contrast, quantum devices naturally model molecular structures according to the laws of quantum physics.

The industry is actively transitioning from laboratory tests to solving complex physical world problems. Key application areas where quantum utility is expected or being tested include:

In Which Areas is Quantum Computing Hard to Achieve?

Businesses are actively preparing for the quantum era: major logistics operators like DHL and financial conglomerates including HSBC and JPMorgan are testing algorithms for process optimization.

However, in the scientific community, these areas are officially recognized as the most complex and farthest from achieving real quantum advantage. The reason is that for most combinatorial problems, such as the classical "traveling salesman problem" or financial portfolio optimization, the best quantum algorithms (QAOA or Grover's algorithm) can only provide quadratic speedup. For a quantum computer to surpass silicon, it will require millions of ideal, fault-tolerant logical qubits.

Other areas where marketing outpaces scientific achievements include:

  • Quantum Machine Learning. For a quantum neural network to process a dataset (e.g., a million images or a terabyte of text), it must be converted from classical binary form "0" and "1" into superposition amplitudes. This requires quantum random access memory (QRAM). The problem is that an effective technology does not yet exist. The process of loading a massive amount of data into qubits takes so long (growing linearly or even superlinearly) that it fundamentally undermines any quantum speedup;
  • Database Management. Modern QPUs operate at frequencies thousands of times lower (kilohertz or megahertz) than CPUs. Due to this colossal difference, Grover's algorithm's quadratic speedup will only start to yield real benefits when the database size becomes truly astronomical. However, such a large database cannot yet be loaded into a quantum computer due to the QRAM issue;
  • Cybersecurity Threats. To break a standard RSA-2048 key, a quantum computer needs about 4,000 logical qubits with FTQC. According to most roadmaps of major projects, this result may be achieved around the 2030s.