Quantum Computers Explained

Decades ago, your mobile phone—or even a microwave oven—would have been the realm of science fiction. And how they would work would have been difficult to explain because general education would not have addressed advanced software and silicon computer chips. Quantum computers fall into this same category. They read like science fiction, and to understand them completely requires training in quantum physics, advanced computation, probabilistic assessments and other disciplines. And like your phone and microwave, they are here today.

The science fiction of the quantum realm

At the scale of atoms and electrons, the rules change from what we are used to seeing around us. In our computers today (and mobile phones), software code, instructions internal to the computer and data are represented by many, many bits with each bit simply being a 1 or a 0. And like the light switch on your wall, we clearly know if the switch is on (1) or off (0).

A quantum computer comprises quantum bits, or qubits, and these are very, very different. When operating they can be in what is called superposition. The idea of superposition may require us to “lean into the science fiction” with a willingness to simply accept the idea: the qubit in superposition is both a 0 and a 1 at the same time. Maybe think of it as a blend of both 0 and 1—but in reality, it is both states at the exact same time.

When you choose to measure the result and get the answer from the qubit (called quantum collapse), it will give you a single answer of either a 1 or a 0—just like the output from a classical computer. The challenge is how do you know the answer is actually correct? That is where probability comes in. Unlike our computers today, where we know the light switch is on or off every time we look, a qubit’s answer comes with odds of being a 1 or 0.

The good news is that those odds are not left to chance—they are engineered. Behind the scenes, every answer a quantum computer could give carries a weight, and those weights behave like waves: they can add together or cancel out. A quantum computation is choreographed so that the paths leading to wrong answers cancel and the paths leading to the right answer reinforce one another. So when you measure, the answer you are most likely to see is the right one. Run the computation a few times and the right answer shows up again and again—that consistency is how you know it is correct.

There is one more piece of science fiction to lean into: entanglement. Qubits can be linked so that measuring one instantly tells you something about its partner, no matter how far apart they are. Einstein himself was skeptical and famously called it “spooky action at a distance”—yet experiment after experiment has shown it is real. Entanglement is what lets qubits work together as a group, and it is why their power grows so quickly. Attempting to write down everything happening inside just 300 entangled qubits would take more numbers than there are atoms in the observable universe. No classical computer could ever store that. So no, a quantum computer does not try every answer at once, as is often said. It does something stranger. It lets all of those possibilities push and pull on one another until the wrong answers fade away (cancel out) and the right one stands out.

Again, like the other advanced technology devices in your life, for quantum computers it is not necessary to understand how they work in detail—that is Phasymmetry’s job. It is important to understand their potential benefits and limitations.

Figure 1. A bit and a qubit.

How we know it works

Quantum mechanics is a century old, among the most precisely tested theories in science, and already inside everyday technology: the transistors in every computer chip, the lasers that carry internet traffic over fiber, MRI scanners, and the atomic clocks that keep GPS accurate. Quantum computers are not hypothetical either. Machines with more than one hundred qubits operate today, and many can be used over the cloud. In 2024 and 2025, research teams showed that quantum error correction behaves as theory predicts: combining more physical qubits into one protected “logical” qubit drives errors down rather than up, a threshold the field had worked toward for about thirty years. In 2025, the Nobel Prize in Physics recognized experiments from the 1980s showing that an electrical circuit large enough to hold in your hand obeys quantum rules, the principle behind superconducting quantum processors. What remains is engineering: building machines large and reliable enough to do useful work that classical computers cannot.

  1. 1920sQuantum mechanics takes shape; a century of precise tests follows.
  2. 1980sA circuit large enough to hold in your hand is shown to obey quantum rules (Nobel Prize in Physics, 2025).
  3. 2024 to 2025Error correction behaves as theory predicts: more physical qubits per logical qubit, fewer errors.
  4. TodayMachines with more than one hundred qubits, many usable over the cloud.
Figure 2. Milestones named on this page.

The potential: how quantum computers will change our world, radically

Today’s quantum computers have physical qubits numbering in the hundreds to low thousands, and error-corrected logical qubits counted in the tens. The most detailed public roadmaps aim for hundreds of logical qubits by the end of this decade and a few thousand in the early 2030s. Look further out, past any roadmap published today, to machines with millions of logical qubits running long computations reliably. At that scale, a quantum computer stops being a specialist instrument and becomes a general resource for the problems classical machines can only approximate.

Figure 3. Logical qubits today, on the most detailed public roadmaps, and at the scale this page looks toward. Orders of magnitude, not forecasts.Source: figures stated on this page.

The first change would be in how we understand and design matter. Nature is quantum, and classical computers can only approximate it. With enough logical qubits, designing a catalyst that makes fertilizer with far less energy, a battery chemistry that stores more and lasts longer, a drug molecule matched to its target before it is synthesized, or a material built atom by atom for strength, weight or conductivity becomes far more a matter of calculation than of trial and error in the lab. Chemistry, materials science, energy and medicine would each be reshaped by that one shift.

The second change would be in decisions with too many moving parts to weigh today: routing fleets, scheduling factories and power grids, building portfolios, and planning supply chains. The gains here vary by problem and are, in theory, often smaller than in chemistry, but at this scale and on problems this large, even modest gains are worth a great deal. Machine learning and financial risk modeling are active areas of research for machines at this scale. One consequence is already being planned for: a machine of this size could break much of the public-key encryption in use today, and governments have published replacement standards that organizations are beginning to adopt. None of this will arrive on a single day. It will arrive the way electricity and the internet did, one industry and one workflow at a time.