First Quantum Computer Explained: Who Built It and When
The phrase "first quantum computer" sounds like it should point to a single machine sitting in a lab somewhere, with a date stamped on its chassis. The actual history is more complicated. Depending on what counts as a quantum computer — a mathematical idea, a two-qubit laboratory experiment, a programmable processor, or a machine sold to a paying customer — the answer shifts by decades. This article breaks the timeline into four distinct milestones: the theoretical concept, the first physical experimental machine, the first universal programmable system, and the first commercial quantum computer.
Milestone One: The Theoretical Concept of Quantum Computing
Before any hardware existed, the idea of quantum computation took shape in theoretical physics. The question that started it all was deceptively simple: can a classical computer efficiently simulate the behavior of quantum systems?
Richard Feynman's 1981 Proposal
In 1981, physicist Richard Feynman delivered a lecture that is now widely regarded as the founding moment of quantum computing. Feynman observed that classical machines appear fundamentally incapable of efficiently simulating quantum mechanical systems. The reason lies in the exponential growth of the state space: describing a quantum system with just a few dozen particles requires more classical memory than any existing computer could hold. Feynman's proposed solution was radical but elegant — to simulate nature, which operates on quantum rules, one would need a computer built on those same quantum rules.
This insight shifted the conversation. Rather than asking whether quantum mechanics could be modeled on classical hardware, Feynman suggested building entirely new machines that exploited quantum phenomena as their core computational resource.
David Deutsch and the Universal Quantum Computer (1985)
Four years later, David Deutsch at Oxford University published a paper that provided the mathematical framework for what he called a universal quantum computer. Deutsch's work went beyond Feynman's conceptual proposal. He described a theoretical machine capable of simulating any physical process and proved that such a machine was physically possible under the laws of quantum mechanics.
Deutsch also introduced the concept of quantum parallelism — the idea that a quantum computer could evaluate multiple computational paths simultaneously through superposition. While the precise mechanism is more subtle than simple parallel processing, Deutsch's paper established that quantum computers could, under certain conditions, solve problems that would be intractable on classical machines.
Peter Shor's Algorithm (1994)
The theoretical foundation gained practical urgency in 1994 when mathematician Peter Shor developed an algorithm for factoring large integers efficiently on a quantum computer. Factoring large numbers is computationally hard for classical machines — a property that underpins widely used encryption systems like RSA. Shor's algorithm demonstrated that a sufficiently powerful quantum computer could break these encryption schemes in polynomial time.
Shor's result transformed quantum computing from an academic curiosity into a field with clear real-world implications for cryptography, security, and computational complexity theory. It also gave experimentalists a concrete target: build a machine that can run Shor's algorithm.
| Year | Researcher | Contribution | Significance |
|---|---|---|---|
| 1981 | Richard Feynman | Proposed building computers based on quantum principles to simulate nature | Identified fundamental limitation of classical simulation; planted the seed for quantum computing |
| 1985 | David Deutsch | Described the universal quantum computer; introduced quantum parallelism | Established mathematical framework proving quantum computers are physically possible |
| 1994 | Peter Shor | Developed Shor's algorithm for integer factorization | First concrete demonstration of quantum advantage for a practically important problem |
Milestone Two: The First Physical Experimental Quantum Computer (1998)
The transition from theory to hardware took over a decade. The first physical implementation of a quantum computer emerged not from a technology company but from academic laboratories working with nuclear magnetic resonance, commonly known as NMR.
The 1998 NMR Experiment
In 1998, researchers led by teams at Oxford and collaborating institutions demonstrated the first working quantum computer using NMR techniques. By applying carefully tuned radio frequency pulses, the team manipulated the nuclear spin states of these atoms, effectively performing quantum logic operations.
The machine operated with just two qubits and executed simple quantum algorithms. While the computational output was modest, the experiment proved that quantum gates — the building blocks of quantum circuits — could be physically realized and controlled in a laboratory setting.
This demonstration answered a critical question: quantum computation was not just a theoretical abstraction. It could be built, operated, and measured using existing laboratory equipment.
Shor's Algorithm on a Seven-Qubit NMR System (2001)
The next significant milestone came in 2001, when researchers implemented Shor's factoring algorithm on an NMR system with seven qubits. The machine successfully factored the number 15 into 3 and 5. From an arithmetic perspective, this result is trivial. From a physics perspective, it was the first experimental demonstration that Shor's algorithm — designed on paper seven years earlier — could actually run on physical hardware.
Why NMR Reached Its Limits
Despite these early successes, NMR-based quantum computers hit a fundamental scaling barrier. The technique relied on measuring signals from large ensembles of identical molecules — typically on the order of 10 to the power of 18 molecules in a liquid sample. As the number of qubits increased, the signal-to-noise ratio degraded exponentially. By the early 2000s, researchers had pushed NMR systems to around 12 qubits, but further scaling became impractical.
The field needed a different physical platform — one that could isolate and control individual quantum systems rather than averaging over ensembles. This requirement drove the development of alternative approaches that would define the next generation of quantum hardware.
Milestone Three: The First Universal Programmable Quantum Computer
Defining the "first universal programmable quantum computer" requires clarifying what the term means. A universal quantum computer is one that can execute arbitrary quantum algorithms — not just a single specialized computation. A programmable system is one where users can define and submit their own quantum circuits, rather than running a fixed experiment designed by the hardware builders.
By this definition, the transition from experimental prototype to universal programmable machine occurred in stages, with different physical platforms reaching this threshold at different times.
Trapped Ion Systems: Early Programmability (2005 Onward)
Trapped ion quantum computers confine individual atoms in electromagnetic fields and use laser pulses to manipulate their quantum states. This approach offered a critical advantage over NMR: each qubit was a single, individually addressable atom rather than an ensemble average.
By 2005, researchers had demonstrated two-qubit logic gates — a threshold that NMR systems struggled to maintain as qubit counts grew. The high fidelity of ion trap operations made them strong candidates for the first genuinely programmable quantum systems.
Trapped ion platforms allowed researchers to define arbitrary sequences of quantum gates and execute them on demand. This flexibility, combined with high gate fidelity, positioned ion traps as one of the earliest platforms meeting the criteria for universal programmable quantum computation.
Superconducting Qubits and Cloud Access (2016)
Superconducting quantum circuits use tiny loops of superconducting material interrupted by Josephson junctions. When cooled to near absolute zero, these circuits exhibit quantum behavior. The approach gained significant momentum because it leverages fabrication techniques similar to those used in conventional semiconductor manufacturing.
A defining moment for superconducting quantum computing came around 2016, when cloud-based access to quantum processors became available to external researchers. For the first time, scientists and developers who had never built a quantum experiment could submit their own quantum circuits to real hardware and receive results. These early cloud-accessible machines typically offered 5 to 20 qubits and were noisy and error-prone, but they represented a meaningful threshold: quantum hardware was no longer confined to the laboratories that built it.
The availability of cloud-accessible quantum processors accelerated algorithm development because researchers could test ideas on physical machines rather than relying solely on classical simulators. This parallel development model — algorithm design advancing alongside hardware improvement — became a characteristic feature of the field's maturation.
What "Universal Programmable" Means in Practice
A universal programmable quantum computer must satisfy several criteria:
- It can execute arbitrary sequences of quantum gates, not just a single pre-programmed algorithm
- Users can define their own quantum circuits and submit them for execution
- The system supports a universal gate set — a collection of quantum gates sufficient to approximate any quantum computation
- The machine produces results that reflect genuine quantum behavior, verifiable through benchmarking against classical simulations for small problem sizes
Both trapped ion and superconducting platforms met these criteria by the mid-2010s, though with different trade-offs. Trapped ion systems offered higher gate fidelity but slower operation speeds. Superconducting systems provided faster gates but shorter coherence times. Neither platform had achieved the error rates required for large-scale fault-tolerant computation, but both demonstrated that universal programmable quantum machines were technically feasible.
| Platform | Approximate Date | Key Limitation |
|---|---|---|
| Trapped Ion | 2005 onward | Slower gate speeds; challenges in scaling to large qubit counts |
| Superconducting Circuits | 2016 (cloud access) | Shorter coherence times (50-200 microseconds); requires cryogenic cooling |
Milestone Four: The First Commercial Quantum Computer (2011)
The first quantum computer sold to an external organization appeared in 2011, when a Canadian company delivered a 128-qubit machine to Lockheed Martin, a major aerospace and defense contractor. This transaction marked the transition of quantum computing from academic research to commercial product — though the nature of the machine and its capabilities sparked significant scientific debate.
The 2011 Delivery: Quantum Annealing vs. Gate-Based Computation
The machine delivered in 2011 used a computational approach called quantum annealing, which differs fundamentally from the gate-based model described in the previous section. Quantum annealing machines are designed to find low-energy states of complex optimization problems. They do not execute arbitrary quantum algorithms or support universal quantum computation.
The distinction matters. A gate-based quantum computer can, in principle, run any quantum algorithm given sufficient qubits and error correction. A quantum annealing machine is specialized for optimization problems — finding the best solution among many possibilities — but cannot execute algorithms like Shor's factoring routine or quantum simulation protocols.
The 2011 delivery was commercially significant: it was the first time an organization paid for and received a quantum computing device. Scientifically, however, the question of whether the machine demonstrated genuine quantum speedup remained contested. Subsequent studies published in peer-reviewed journals found evidence of quantum behavior within the device but no conclusive computational advantage over well-optimized classical algorithms for the problems tested.
Subsequent Commercial Developments
Following the 2011 delivery, the commercial quantum computing landscape diversified. Cloud-based quantum processors became accessible to researchers around 2016, offering 5 to 20 qubits through internet-connected platforms. These systems, while still noisy, allowed a broader community to experiment with real quantum hardware rather than theoretical models.
In 2019, a landmark experiment demonstrated what the research community termed "quantum supremacy" — the completion of a specific computational task that would take a classical supercomputer an impractical amount of time to reproduce. The task itself had no known practical application; it was designed to highlight raw computational capability. The result confirmed that quantum machines had entered a regime where classical computers could no longer easily simulate their behavior, though the practical significance of this threshold remains a subject of ongoing discussion.
| Year | Milestone | Qubits | Type | Significance |
|---|---|---|---|---|
| 2011 | First commercial delivery to Lockheed Martin | 128 | Quantum annealing | First sale of a quantum computer to an external organization; sparked scientific debate about quantum speedup |
| 2016 | Cloud-accessible quantum processors | 5-20 | Gate-based (superconducting) | Opened quantum hardware to external researchers; accelerated algorithm development |
| 2019 | Quantum supremacy demonstration | 50+ | Gate-based (superconducting) | First task completed by quantum machine that classical supercomputers could not practically reproduce |
Comparing the Four Milestones
The four milestones described above represent fundamentally different achievements. Understanding the distinction between them is essential for evaluating claims about "the first quantum computer" in both scientific literature and popular media.
| Milestone | Date | Nature | Practical Utility |
|---|---|---|---|
| Theoretical concept | 1981-1994 | Mathematical framework and algorithms | None directly; established feasibility and motivation for building physical machines |
| First experimental machine | 1998 | Two-qubit NMR system executing simple algorithms | Proof of principle; demonstrated that quantum gates can be physically realized |
| First universal programmable system | 2005-2016 | Trapped ion and superconducting platforms supporting arbitrary gate sequences | Enabled external researchers to run custom quantum circuits; accelerated algorithm development |
| First commercial quantum computer | 2011 | 128-qubit quantum annealing machine | Specialized for optimization problems; not universal; practical advantage over classical methods remains debated |
Each milestone built on the ones before it. The theoretical work of the 1980s and early 1990s defined what a quantum computer should do. The NMR experiments of the late 1990s showed that it could be done at all. The trapped ion and superconducting platforms of the 2000s and 2010s made it programmable. The commercial deliveries of the 2010s made it available to organizations outside the research community.
Frequently Asked Questions
When was the first quantum computer built?
The answer depends on the definition. The first theoretical proposal came from Richard Feynman in 1981. The first physical experimental quantum computer was demonstrated in 1998 using nuclear magnetic resonance on a chloroform molecule with two qubits. The first commercial quantum computer was delivered in 2011.
Who built the first quantum computer?
The theoretical foundation was laid by Richard Feynman (1981), David Deutsch (1985), and Peter Shor (1994). The first physical implementation was achieved by research teams at Oxford and collaborating institutions in 1998, using NMR techniques. The first commercial machine was delivered by a Canadian company in 2011.
How many qubits did the first quantum computer have?
The first experimental quantum computer in 1998 operated with 2 qubits. The first commercial quantum computer in 2011 had 128 qubits but used quantum annealing rather than the gate-based model. Cloud-accessible gate-based systems in 2016 offered 5 to 20 qubits.