In this article, we explore the companies at the forefront of quantum computing and their advancements in qubit technology. These companies are categorized based on their unique approaches to quantum computing, as detailed in the previous video, “The Map of Quantum Computing.” If you haven’t seen it, you might find it helpful to watch.
In the realm of quantum computing, qubit counts are often highlighted. Yellow numbers denote universal quantum computing qubits, while orange numbers represent non-universal quantum computers or simulators. Although these technologies are intriguing, they cannot be directly compared to universal quantum computers. Additionally, several non-hardware companies and software packages are included for a comprehensive view.
Some companies do not disclose their qubit counts, as this information is not yet public. Many startups are diligently working on their machines but have not released specific details. The field of quantum computing is rapidly evolving, and this snapshot reflects the state of the industry as of January 2022. However, it may soon become outdated.
When assessing which company is building the best quantum computers, qubit counts alone are insufficient. A higher number of qubits does not necessarily equate to better performance. For instance, a thousand low-quality, high-noise qubits may be less effective than ten high-quality, low-noise qubits. Moreover, not all qubit counts satisfy the criteria for a fully functional quantum computer, known as DiVincenzo’s criteria. Understanding these criteria is crucial when evaluating different systems.
Future press releases and news reports should provide more than just qubit counts. It’s essential to understand the technological advancements of a company’s quantum computers, considering factors like error rates, crosstalk, and connectivity.
Fortunately, we have a benchmarking tool called quantum volume, developed by IBM, to compare different quantum systems. Quantum volume measures the effectiveness of a quantum computer, offering an objective comparison across various architectures. It helps set measurable goals for future advancements.
Quantum volume is defined as the largest square circuit a quantum computer can solve. A circuit comprises a series of instructions for the qubits. For example, a circuit with four qubits and four time steps corresponds to a quantum volume of 16. Adding one more qubit and one more time step doubles the quantum volume.
Currently, the highest quantum volume is achieved by a trapped ion quantum computer from Continuum, a company formed by merging Honeywell and Cambridge Quantum. Their machine has 12 qubits and a quantum volume of 2048. In contrast, IBM’s 127-qubit superconducting quantum computer has a quantum volume of 128. This demonstrates how a machine with fewer qubits can outperform one with more qubits using this metric.
However, not all companies have published their quantum volume numbers, limiting our ability to compare all quantum computers using this metric. While quantum volume is valuable, additional metrics will be needed in the future. IBM has proposed another metric called CLOPS to address throughput issues.
According to published quantum volume data, Continuum’s system model H12 currently leads the field. However, this information may soon change as the industry progresses.
It’s important to note that no quantum computers today can solve real-world problems more efficiently than classical computers. Here’s a roadmap of challenges that need to be addressed to develop a useful quantum computer, based on insights from the Google Quantum Technology blog:
Currently, companies are focusing on the first two tasks, each presenting significant engineering challenges.
Many companies have set ambitious goals. IBM plans to have a 433-qubit machine in 2022 and over a thousand qubits by 2023, with aspirations for up to a million qubits by 2026. Google aims for a useful error-corrected quantum computer by the end of the decade, targeting around a million physical qubits by 2030. D-Wave is working on a quantum annealing chip with 5,700 qubits and plans to build over 7,000 qubits by 2024. Other companies, like PsiQuantum and ColdQuanta, have their own ambitious predictions for the coming years.
While this overview does not cover all companies, it highlights some of the most notable predictions. If these projections hold true, we could see useful general-purpose quantum computers in about eight years.
In the future, it is hoped that companies will not only report qubit counts but also confirm compliance with DiVincenzo’s criteria and publish quantum volume results or other relevant metrics. This information is essential for those outside the industry to critically assess their claims.
Research the current qubit counts of leading quantum computing companies. Compare and contrast their approaches, and discuss how these counts impact their technological capabilities. Present your findings in a short report.
Investigate the concept of quantum volume and its significance in evaluating quantum computers. Create a presentation explaining how quantum volume is calculated and why it is a more comprehensive metric than qubit count alone.
Organize a workshop to delve into DiVincenzo’s criteria for a fully functional quantum computer. Discuss each criterion in detail and evaluate how different companies meet these standards. Prepare a group presentation summarizing your insights.
Engage in a debate about the future projections of quantum computing companies. Analyze their goals and timelines, and argue whether these projections are realistic. Support your arguments with data and trends from the industry.
Create a roadmap outlining the key challenges and milestones in developing a useful quantum computer. Use insights from the article and additional research to highlight the steps companies need to take to achieve their goals.
Here is a sanitized version of the provided YouTube transcript:
—
In this video, I will discuss the companies currently developing quantum computers and their highest qubit counts. I’ve organized them based on their approaches to quantum computing, which I explained in my previous video, “The Map of Quantum Computing.” If you haven’t seen that yet, you might want to check it out.
The yellow numbers represent universal quantum computing qubits, while the orange numbers indicate various non-universal quantum computers or simulators. These are interesting technologies, but they cannot be directly compared to universal quantum computers. For completeness, I’ve also included several non-hardware companies and software packages.
You may have noticed that some companies do not have a qubit count listed. This is because that information is not publicly available yet. Many startups are actively working on their machines but have not released details. I apologize if I made any mistakes or left anyone out; the field has expanded significantly in recent years, and I had to set a limit on my research. This snapshot reflects the state of the field as of January 2022, but it may become outdated quickly.
When evaluating which company is building the best quantum computers, we cannot rely solely on qubit counts. The number of qubits alone is not a reliable measure for comparing different quantum systems. For instance, a thousand low-quality, high-noise qubits may perform worse than ten high-quality, low-noise qubits. Additionally, not all qubit counts meet the full criteria to be considered a fully functional quantum computer, known as DiVincenzo’s criteria. It can be challenging to determine which companies meet these criteria, but it’s important to be aware of them.
In the future, when press releases and news reports are published, we should seek more information beyond just qubit counts. We need to understand how advanced a company’s quantum computers are, as many factors influence their performance, including error rates, crosstalk, and connectivity.
Fortunately, we can compare different systems using a benchmarking scheme called quantum volume, developed by IBM. Quantum volume is a metric that indicates how effective a quantum computer is and provides an objective way to compare machines, regardless of their underlying architecture. This metric helps set measurable goals for the future.
Quantum volume is defined as the largest square circuit that a quantum computer can solve. A circuit consists of a list of instructions for the qubits. For example, if a circuit involves four qubits and four time steps, it corresponds to a quantum volume of 16. Each time you add one more qubit and one more time step, the quantum volume doubles.
Currently, the machine with the highest quantum volume is a trapped ion quantum computer by Continuum, a company formed from the merger of Honeywell and Cambridge Quantum. Their machine has 12 qubits and a quantum volume of 2048. In comparison, IBM’s 127-qubit superconducting quantum computer has a quantum volume of 128. This illustrates how a machine with fewer qubits can outperform one with more qubits under this metric.
However, to my knowledge, no other companies have published their quantum volume numbers, so we cannot yet compare all quantum computers using this metric. While quantum volume is a useful metric, we will need additional metrics in the future. For instance, IBM has proposed another metric called CLOPS, which addresses throughput issues.
To summarize, according to the published quantum volume, Continuum’s system model H12 is currently the best quantum computer. However, this information may quickly become outdated.
Looking ahead, it’s important to note that none of the quantum computers available today can solve real-world problems better than existing classical computers. Here’s a roadmap of the challenges that need to be addressed to create a useful quantum computer, based on information from the Google Quantum Technology blog:
1. Implement error correction by combining several physical qubits to create a single error-corrected qubit, known as a logical qubit.
2. Demonstrate that adding more physical qubits improves error correction.
3. Create a logical qubit that can maintain coherence indefinitely.
4. Develop two error-corrected logical qubits to form a quantum transistor for two-qubit operations.
5. Scale up to hundreds or thousands of these qubits to build a full-scale error-corrected quantum computer.
Currently, companies are focusing on the first two tasks, each of which presents significant engineering challenges.
As for timelines, many companies have set ambitious goals. IBM plans to have a 433-qubit machine in 2022 and over a thousand qubits by 2023, with aspirations for up to a million qubits by 2026. Google aims for a useful error-corrected quantum computer by the end of the decade, targeting around a million physical qubits by 2030. D-Wave is working on a quantum annealing chip with 5,700 qubits and plans to build over 7,000 qubits by 2024. Other companies, like PsiQuantum and ColdQuanta, have their own ambitious predictions for the coming years.
While this overview does not cover all companies, it highlights some of the most notable predictions. If these projections hold true, we could see useful general-purpose quantum computers in about eight years.
In the future, I hope companies will not only report qubit counts but also confirm compliance with DiVincenzo’s criteria and publish quantum volume results or other relevant metrics. This information is essential for those of us outside the industry to critically assess their claims.
Thank you for watching! If you enjoyed this video, please like and subscribe.
—
This version removes informal language and any inappropriate expressions while maintaining the original content’s meaning and structure.
Quantum – Quantum refers to the smallest possible discrete unit of any physical property, often used in the context of quantum mechanics, which studies the behavior of matter and energy at the atomic and subatomic levels. – In quantum mechanics, particles can exist in multiple states at once, a phenomenon known as superposition.
Computing – Computing is the process of utilizing computer technology to complete a given goal-oriented task, often involving the processing of data and execution of algorithms. – Quantum computing leverages the principles of quantum mechanics to process information in ways that classical computers cannot.
Qubit – A qubit is the basic unit of quantum information, analogous to a bit in classical computing, but capable of representing both 0 and 1 simultaneously due to superposition. – The entanglement of qubits allows quantum computers to solve complex problems much faster than traditional computers.
Volume – In physics and engineering, volume refers to the amount of three-dimensional space occupied by an object or substance. – The volume of the gas in the chamber was measured to determine its pressure and temperature relationship.
Error – An error in physics and engineering is the difference between a measured value and the true value, often due to limitations in measurement precision or external factors. – Reducing error in experimental measurements is crucial for obtaining accurate and reliable results.
Correction – Correction in the context of physics and engineering refers to the process of adjusting calculations or measurements to account for known errors or biases. – Error correction codes are essential in quantum computing to maintain the integrity of qubit states.
Performance – Performance in engineering and computing refers to the effectiveness and efficiency with which a system or component operates, often measured against specific criteria or benchmarks. – The performance of the new processor was evaluated based on its speed and power consumption.
Technology – Technology in the context of physics and engineering refers to the application of scientific knowledge for practical purposes, especially in industry and the development of devices or systems. – Advances in semiconductor technology have significantly increased the computational power of modern devices.
Metrics – Metrics are standard measures used to assess the performance, quality, or efficiency of a system or process in engineering and computing. – Key performance metrics for the new software include processing speed and error rate.
Engineering – Engineering is the application of scientific and mathematical principles to design, build, and maintain structures, machines, and systems. – Civil engineering projects require careful planning and execution to ensure the safety and durability of structures.
