John Martinis is fundamentally a hands-on experimental physicist. He thrives in the practical realities of laboratory work, a stark contrast to the theoretical neatness found in textbooks. Yet, his contributions are so significant that any comprehensive history of quantum computing would be incomplete without his inclusion. Martinis played a central role in two of the field’s most groundbreaking achievements and remains dedicated to pursuing the next frontier.
Pioneering Macroscopic Quantum Phenomena
His early career in the 1980s saw Martinis and his colleagues conduct a series of key experiments aimed at exploring the boundaries of known quantum effects. This work, which earned him a Nobel Prize the previous year, began during his doctoral studies at the University of California, Berkeley. At the time, the scientific community acknowledged that subatomic particles adhered to quantum principles. The crucial, unanswered question was whether these quantum behaviors could extend to larger, more macroscopic scales.
Martinis and his research team constructed and investigated circuits composed of a combination of superconductors and insulators. Their findings revealed that many charged particles within these circuits behaved as a unified quantum entity. This phenomenon, known as macroscopic quantumness, established the groundwork for developing some of today’s most powerful quantum computers, including those advanced by major corporations like IBM and Google. Indeed, Martinis’s research initiated the trend of technology leaders adopting superconducting circuits for fabricating quantum bits, or qubits—the most prevalent type of qubit currently in use worldwide.
Achieving Quantum Supremacy
The second transformative contribution from Martinis came when he led the Google research team responsible for building the quantum computer that first demonstrated “quantum supremacy.” For almost five years, this machine stood alone as the only device globally, whether quantum or classical, capable of validating the results of a random quantum circuit. Ultimately, classical computers were developed that could outperform it.
Founding QoLab and Pursuing Practical Quantum Computing
Now approaching his 70th birthday, Martinis believes he is poised to achieve another historic milestone with superconducting qubits. In 2024, he co-founded QoLab, a quantum computing company. Martinis states that QoLab will adopt a fundamentally new strategy in its pursuit of what the entire field has been striving for: genuinely practical quantum computers.
Early Insights and Technological Evolution
The Genesis of a Technological Vision
Karmela Padavic-Callaghan: You achieved significant groundbreaking work early in your career. When did you first recognize that your experiments held the potential to spawn a new technology?
John Martinis: There was an open question regarding whether a macroscopic variable could elude quantum mechanics. As a young researcher new to quantum mechanics, it seemed imperative to test this. Perhaps an older researcher might have simply assumed quantum mechanics would apply universally. However, for a young student, it presented itself as a compelling experiment for a fundamental test of quantum mechanics.
We initially set up a very basic, high-speed experiment using the technology available at the time. When we examined the data, the experiment was a complete failure. Yet, we were able to learn from the failure quickly, which minimized its impact. Ultimately, the experiment demanded a solid understanding of microwave engineering and noise characteristics, involving numerous technical challenges. Success followed relatively soon after.
Over the subsequent decade, we refined this experimental approach and began constructing quantum devices. Concurrently, the theoretical framework of quantum computing advanced considerably, particularly with algorithms like Shor’s, which is crucial for breaking cryptography by factoring large numbers. Error-correction algorithms followed shortly thereafter. These developments provided a robust foundation for the field, enabling researchers to envision the construction of actual quantum systems. This progress subsequently led to increased funding.
The Impact of Funding on Research and Development
Karmela Padavic-Callaghan: How did funding influence research and, consequently, the technology?
John Martinis: The landscape has transformed dramatically since the 1980s. Back then, the ability to manipulate and measure a single quantum system had not even been thoroughly tested. It is remarkable to observe the progress made over the past four decades. Quantum computing has truly evolved into a massive field! The most gratifying aspect is the substantial number of physicists now employed to study the quantum mechanics of these superconducting systems and to develop quantum computers.
Leveraging Long-Term Experience
Navigating the Future Through Past Expertise
Karmela Padavic-Callaghan: You were involved in the nascent stages of quantum computing. How does that experience inform your understanding of the field’s current trajectory?
John Martinis: Having been immersed in the field throughout its development, I possess a deep understanding of its fundamental physics. I was responsible for building the initial microwave electronics for quantum devices within our group at the University of California, Santa Barbara. Later, at Google, I personally constructed cryostats – essential for maintaining the extremely low temperatures required for superconducting quantum computers to function. My involvement extended to the fabrication of every component. I believe many individuals, without this comprehensive experience, might harbor an optimistic outlook on continued progress. I, however, am acutely aware of the existing challenges. Building a highly complex computing system is an immense undertaking in systems engineering. My advantage lies in my thorough grasp of the underlying physics governing all aspects.
Rethinking Quantum Computer Hardware for Practicality
Karmela Padavic-Callaghan: How must quantum computing hardware evolve to render quantum computers practical and useful? What specific advancements do you anticipate will trigger the next major breakthrough?
John Martinis: Following my departure from Google, I began to view quantum computers as integrated systems, fundamentally re-evaluating the essential requirements for building and improving them. QoLab’s strategy is rooted in these reconsiderations, incorporating significant changes in qubit manufacturing techniques and the overall assembly of the system, particularly concerning the wiring.
We recognized that developing reliable quantum computers cost-effectively necessitates a completely different approach to their construction. This concept is challenging to articulate and for people to fully grasp. Despite encountering considerable resistance and skepticism, my extensive experience in physics over many decades suggests that this indicates we are on the right track.
Addressing the Million-Qubit Challenge
Karmela Padavic-Callaghan: The development of error-free, truly useful quantum computers is often linked to the need for millions of qubits. How do you envision achieving this scale?
John Martinis: Our primary focus for significant disruption lies in manufacturing, specifically in the fabrication of quantum chips, which is inherently the most demanding aspect. Observing the methods employed by major players like Google, IBM, Amazon, and numerous other companies reveals a reliance on manufacturing techniques dating back, I would estimate, to the 1950s or 1960s. It is difficult to identify another industry currently using such outdated methods for producing intricate circuits. Therefore, our perspective is that achieving reliable systems with a million qubits requires adopting alternative strategies.
We are exceedingly optimistic about our capacity to fundamentally alter the manufacturing process for these devices. Furthermore, we have developed an architectural design for the chips that eliminates the need for external wiring. Images of superconducting quantum computers often display a complex network of wires and microwave components. Our goal is to integrate these elements onto a chip and enable scalable production. The primary obstacle for superconducting qubits is the wiring complexity, and we are actively working to resolve this.
The Race for Practical Quantum Computing
Karmela Padavic-Callaghan: Do you foresee a clear leader emerging in the pursuit of a practical quantum computer within the next five years?
John Martinis: Various approaches are being explored for building quantum computers. Given the stringent systems engineering constraints, I believe that tackling this challenge through multiple diverse methods is beneficial. I see value in the funding of numerous distinct ideas, as this increases the probability of a breakthrough. However, when considering these constraints, and there are many, I generally find that many current projects exhibit a degree of naivete regarding the actual requirements, such as cost management or scaled device production. Conversely, I am certain that many research teams possess undisclosed strategies for overcoming their design challenges.
QoLab’s business model, I believe, is somewhat distinct, perhaps even unique, in its embrace of collaboration, as we recognize the necessity of collective expertise. We are partnering with hardware manufacturers who possess extensive experience in scaling production and executing sophisticated manufacturing processes.
The First Application for a Hypothetical Quantum Computer
Karmela Padavic-Callaghan: If you were presented with a highly advanced, error-proof quantum computer tomorrow, what would be your immediate priority?
John Martinis: I am particularly interested in leveraging quantum computers to address challenges in quantum chemistry and quantum materials. Recent publications highlight their use in extracting more valuable information from nuclear magnetic resonance (NMR) experiments in chemistry, an application I find highly compelling. This quantum problem presents significant difficulties for classical supercomputers due to the inherent complexities of quantum mechanics. However, it is inherently solvable with a quantum computer, essentially mapping one quantum problem onto another. This prospect is exciting, partly because I prefer having well-defined methods for device construction, and researchers have developed precise algorithms for applications like enhancing NMR.
Many might consider tackling optimization problems or quantum artificial intelligence. For me, those lean more towards an experimental approach. The theoretical underpinnings for materials and chemistry applications are far more established. We have a clear understanding of the required quantum computer size. I believe such a machine, in terms of both scale and processing speed, is attainable.
The Timeline for Quantum Chemistry Applications
Karmela Padavic-Callaghan: Could 2026 mark the year we begin utilizing quantum computers for chemistry?
John Martinis: Understanding the chemical properties of a molecule is an inherently quantum-related problem, making quantum computers a suitable tool for this task—and we might see significant progress in this area starting in 2026.
Bridging Theory and Physical Implementation
Karmela Padavic-Callaghan: While the mathematical underpinnings for some quantum computer applications were established over 30 years ago, why haven’t they materialized yet?
John Martinis: It is possible to abstract the behavior of a qubit and conceptualize the construction of a quantum computer, which is advantageous as it allows computer scientists, mathematicians, and theorists to engage with the problem. However, the actual challenge lies in the fact that real qubits are susceptible to noise sources, such as ambient heat from external wiring or impurities within the qubit material itself—these are physical impediments. Many prominent quantum computing initiatives are led by theorists, which is valuable; however, the real-world system is considerably more complex, as is the effort required to develop functional hardware.
Within the research group of my graduate advisor, John Clarke, I was trained to understand noise. This background proved exceedingly beneficial for myself and my colleagues, as we approached qubits from a distinctly physical perspective, focusing on eliminating the physical noise mechanisms that compromise chip reliability. This was evident in the quantum supremacy experiment; some noise originates from the presence of “two-level states” within the device, and operations are designed to circumvent them. While functionality can be achieved, it is a remarkably difficult and complex process, hindering scalability. My hope is that we can now eliminate or significantly reduce this effect. Addressing this requires delving into the intricate details of qubit design.
The fundamental issue is the simultaneous need for both robust hardware and well-defined application concepts. I believe substantial improvements in hardware across the entire field are still necessary. This is precisely where my current focus lies.