Peter Zoller calls the rush for the first industrial quantum computer a modern Everest expedition. It is obsessive. Terrifyingly so. When you are climbing, all you see is who is ahead of you. When you reach the peak? You ask yourself why you bothered in the first place.
Back in 1995, Zoller and Ignacio Cirac were way ahead of the curve. They, at the University of Colorado, mapped out blueprints using trapped ions. These aren’t regular bits. They are qubits. Superposed entities that sit somewhere between zero and one, holding every state in between. They weren’t even the first to think of it. Paul Benioff and Richard Feyman had guessed a decade earlier that harnessing quantum weirdness could beat classical machines. But Zoller and Cirac provided the how.
Today? It’s a race. IBM and Atom Computing are leading with processors packing over 1,000 qbits. Caltech researchers reported a 6,000-qubit array last year.
“It’s exciting,” says Nobel laureate John Martinis. “People are fielding computers with hundreds, thousands of qubits.”
But let’s check the score.
In 2019, Google claimed “quantum advantage.” Their 53-qubit Sycamore chip did a calculation in 20 seconds. They said a classical supercomputer would need 10,00 years. IBM disagreed. They said it would take two and a half days. The number got lower. And lower. But here is the kicker. The calculation was useless. Pure academic flexing. Kihwan Kim called it a milestone, sure, but not a breakthrough for anything practical.
We are waiting. Years away from practical application. Experts want a million qubits. Robust ones. Error-free. In March, a group showed IBM’s 50-qubit Heron processor could predict neutron scattering for an antiferromagnetic crystal. Accurately. But classical computers do it faster. And more accurately.
So. What is this all for?
“We can now simulate things like superductivity, artificial photosynthesis and small drag designs.”
—Michelle Simmons, Silicon Quatum Computing
Martinis worries about scaling. He notes that million-qubit machines might never exist. The proof, he says, is building them. And seeing they work.
The Encryption Nightmare
The most famous fear is RSA encryption. Banks, crypto, emails. It all rests on big prime numbers. Easy to multiply. Impossible to divide. Or so we thought.
Quantum computers don’t care about impossible. Peter Shor proposed an algorithm in 1994 that exploits qubit entanglement to crack RSA. It checks every possible path at once.
Here is the shocker. It won’t take a million qbits.
An Australian team from Iceberg Quantu calculated in Febuary that fewer than 100k qbits, with error correction, might crack it. Google jumped into action, promising migration by 2027. Artur Ekert, usually skeptical of the hype, is now concerned. “The threat is not mumbo jumbo.” John Martinis thinks five to ten years is a realistic timeline for a breach.
NIST released new post-quantum cryptography standards in 2026. Zoller says this defuses the panic. We can switch protocols before hackers arrive. Shor’s algorithm remains a historical landmark. More important for inspiration than destruction.
Ekert remains anxious. Last year a Chinese scientist claimed to have broken the new lattice-based standards. The claim was wrong. The global community had to spend weeks proving it. It felt too close. What if the next one isn’t wrong?
Simulating The Unseen
Physics needs quantum help. Feyman said we cannot understand nature without building it. Michelle Simmons agrees. Classical computers choke on multi-particle interactions. The data grows exponentially. You run out of memory.
Quantum simulators handle complexity naturally.
Progress is happening. Kim’s team kept trapped-ion coherence for over an hour. Last year two independent groups simulated string breaking. This is the Standard Model’s elastic bond between quarks. Pull it, and it snaps, creating matter and antimater. Pedram Rousharn ran this on Sycamore. He called it a picture of a string snapping. Theory from the 197os made visible.
González-Cuadra worked on an analog version. Neutral atoms in a lattice. They simulated fluctuations in two dimensions. Digital is general. Analog is specialized. Both showed new things.
Will these replace particle colliders? No. But they sharpen theories. They give experimenters something to test. And yes, the verification problem looms large. Zoller recently posted a paper on arXiv about verifying the output of quantum simulators. How do you know the black box isn’t lying to you?
Materials And Medicine
Here is where the money lies. Or the future lies. Zoller wants to move from discovery mode to active design mode. We feed the quantum computer our requirements. It spits out a molecular recipe. Better batteries. New drugs. Without rare earth metals.
“It’s about efficiency,” says Martinis. Saving even a few percent in cost or performance pays off in billions. Ecological shifts are possible.
The holy grail is room-temperature supercondu. No resistance. No cooling required.
Currently superconductivity needs absolute zero-ish conditions. Some materials work at higher temps. We don’t know why. Henrik Dreyer at Quantinuum uses a 98-qbit chip to model cuprate supercondutors. Under laser fire electrons pair up briefly. They flow without resistance.
“The ultimate question is: Can we engineer it for longer?” asks Dreyer. An hour? Ten days? We are off by an order of magnitude. Error rates are currently one in thousand. We need one in million.
Simmons’ team in Australia continues pushing forward with their silicon-based architecture.
So.
When does this become real? Sergio Boixo predicts practical apps in five years. Ekert worries about the cracks in our digital fortress. Zoller looks at the mountain top and sees only the view from the peak. The climb has just begun.
Will you remember when the old world broke? Or only when the new one started to hum?
