What a Quantum Computer Actually Is
A quantum computer is not a faster laptop. It is a different kind of machine. It trades certainty for probability, and it is very good at a very small number of things. One of those things is a threat to the math that protects payments. Here is what one actually is, and why people who guard money are watching it closely.
The machine you already know
Every computer you use runs on bits. A bit is a switch, set to 0 or 1, and at any moment it is firmly one of them. Flip billions a second and you get fast, exact, repeatable work. The same input gives the same answer every time. Hold on to that word, exact. A quantum computer gives that up on purpose.
The qubit, and superposition
A quantum computer runs on qubits. A qubit can be 0, or 1, or a weighted mix of both at the same time. That mix is called superposition. A bit is a light switch. A qubit is a dimmer that can rest anywhere between off and on, leaning further toward one side or the other.
While a qubit sits in that mix it has not picked an answer yet. The lean is described by numbers called amplitudes, and those amplitudes are the whole game. One qubit is a single lean. The power comes from stacking them. Each qubit you add doubles how many combinations the machine holds at once.
Twenty qubits already hold a million combinations together. Sixty hold more combinations than there are grains of sand on Earth, all carried at the same time.
You only get one look
Here is the twist that makes quantum computing strange. You cannot read the mix. The moment you measure a qubit it drops to a plain 0 or a plain 1, and which one you get is down to the odds set by those amplitudes. Set the lean below and measure it. One reading tells you almost nothing. Keep measuring and the results settle toward the odds you set. That is what probabilistic means in practice.
Interactive · measure a qubit
So a classical computer walks one path to one sure answer. A quantum computer holds many at once and then pays out by probability. The craft is making the answer you want almost certain before you look.
The trick that makes it useful
If qubits just gave random results, none of this would matter. The thing that makes it work is interference. Amplitudes behave like waves. Line two up in step and they reinforce into something bigger. Put them out of step and they cancel to nothing. A quantum algorithm is built so the wrong answers cancel out and the right one adds up, so that when you finally measure, the good answer is the one almost sure to appear.
There is a partner trick called entanglement, where two qubits link so tightly that they act as one and reading one tells you about the other. Interference and entanglement are the levers an ordinary computer does not have.
What it is actually good for
A quantum computer is not faster at everything. It only wins where someone has found a way to use interference to reach the answer. That list is short. Breaking certain hard math problems, like factoring huge numbers. Searching enormous spaces, with a smaller speedup. Simulating molecules and materials, which ordinary computers are genuinely bad at, and which could reshape medicine, batteries, and chemistry. For your email and your spreadsheets it offers nothing. Think rare specialist, not replacement.
Where things stand today
Progress is real and fast, but the headline qubit numbers hide a catch worth understanding. There are two ways to count qubits, and they are miles apart.
Physical qubits are the raw hardware, and they are error-prone. By 2026 the biggest machines carry impressive counts. Neutral-atom systems from companies like Infleqtion and Atom Computing have passed a thousand physical qubits. IBM ran a chip above a thousand and then deliberately stepped back to smaller, cleaner ones like its 156-qubit Heron. Google's Willow holds 105. Trapped-ion machines from IonQ and Quantinuum sit in the tens to low hundreds, but with very high quality.
Logical qubits are the ones that matter. A logical qubit is a single reliable qubit built from many physical ones stitched together by error correction. Shor's algorithm needs these, not the raw kind. Here the numbers are tiny. The current lab records sit around ninety logical qubits, and most of those are fragile or counted under favorable conditions. The real working number is dozens, not thousands.
What is coming is a race to close that gap. IBM has a public roadmap toward a fault-tolerant machine with a couple of hundred logical qubits around 2029, and modular systems reaching a hundred thousand physical qubits. Google is aiming for a fault-tolerant computer before 2030. IonQ and Quantinuum are pushing toward a thousand ion qubits with error correction built in. Most serious forecasts put the first genuinely useful, fault-tolerant machines somewhere in the 2028 to 2032 window. Close enough to plan around, not here yet.
Why payments should care
One of those few winning jobs is the problem that holds up modern encryption. The security behind card payments, and most of the internet, rests on a bargain. Scrambling a message is easy. Unscrambling it without the key is so slow that no classical computer could finish in a human lifetime. Factoring the enormous numbers behind RSA, and the related math behind elliptic-curve keys, is exactly that kind of slow problem.
Today
The keys protecting card data and secure connections hold because breaking the math would take a classical machine longer than the age of the universe.
A large quantum computer
An algorithm called Shor's is built to factor those same numbers quickly. Big enough, it unpicks the keys that keep payments and traffic private.
That machine does not exist at the needed scale yet. As we just saw, today's best systems manage dozens of logical qubits, and unpicking RSA is estimated to need thousands of them, which means millions of physical qubits. The gap is large, but it has been shrinking fast. The most-cited estimate fell from around twenty million physical qubits a few years ago to under a million more recently, as the methods improved. The worry is not only the day the machine arrives. It is harvest now, decrypt later. An attacker can store encrypted traffic today and open it the moment a capable machine exists. That is why the standards bodies have already published post-quantum algorithms, and why the payments world has started moving its keys to math a quantum computer cannot unwind.
So a quantum computer is not a better computer. It is a different one. It trades certainty for probability and leans on interference to surface the one answer that counts. It does almost nothing well. One of the few things it may do well is quietly undo the math beneath payments, which is why the ground under encryption is already shifting.