Your system is set to reduce motion, so the core is paused. This screen turns slowly and has a pale white flash about every thirty minutes, when the core vents.
This page is the control room of an imaginary quantum computer, the kind a science-fiction film would show. In the middle, a lattice of glowing qubits turns slowly inside three containment rings. About once a minute a calculation cycle charges the core, light runs along the links between the qubits, and the cycle ends in a soft flash as the qubits are measured and a result appears. Every few minutes a decoherence spike sets off a containment alert and the rings tighten. About every half hour the core runs away: a one-minute overload countdown, an emergency vent in a white flash, half a minute of darkness, and a cold restart that brings the qubits back one by one. It is calm for most of its ten hours. Click ⛶ FULLSCREEN and it fills the screen.
No. It is an animation, and the bar under it says SIMULATION. The facility, the processor called QPU-7, the calculations and their results, the temperatures, and every line of the system log are invented and drawn in your browser. Nothing is computed on any quantum hardware, and the page makes no connections while it runs.
It is also a science-fiction picture, not a portrait of a real machine. Real quantum computers have no glowing core, no containment rings, and nothing that can overload, vent plasma, or explode. Some of the labels are borrowed from real quantum computing, such as the dilution fridge, the surface code, T1 and T2, and the names of algorithms, but the way they behave here is invented for the drama. The section on real quantum computers below explains what those words mean and what the real machines are like.
Three kinds of event repeat through the ten hours, layered over a slow, steady background.
Moments worth seeing in the first part of the cycle (seek with the bar or use ?t=):
| Time | Seconds | What happens |
|---|---|---|
| 00:00:03 | 3 | Cycle 4,410 starts: phase estimation (QPE) |
| 00:00:57 | 57 | Measurement: the qubits turn amber and cyan and the core glows; the result shows at 1:00 |
| 00:04:00 | 240 | First decoherence alert, on ring R3, for about 47 seconds |
| 00:13:00 | 780 | First core overload: the one-minute countdown starts |
| 00:14:00 | 840 | Emergency vent with a white flash; the core goes dark at 14:04 |
| 00:14:16 | 856 | Cold restart: the qubits light up one by one |
| 00:14:40 | 880 | Core online; calculation cycles resume at 14:57 |
| 00:42:36 | 2557 | Second overload |
Over the full ten-hour cycle there are 520 calculation cycles, 62 decoherence alerts, and 20 overloads, spaced 28 to 33 minutes apart.
?t= and the number of seconds to the address, for example /quantum-computer?t=780 for the first overload. Minutes and hours also work: ?t=13m or ?t=1h13m.The page draws at the full resolution of your screen, up to 4K. On a slow computer it lowers the resolution by itself to keep the motion smooth.
Background screens for science-fiction films, series, and music videos with a lab, a spaceship, or a research facility in the story; ambient visuals for ambient, electronic, and lo-fi music streams; a "the machine is thinking" screen for escape rooms, tabletop role-playing sessions, and live-action games; and a calm, slowly changing display for a science event, a classroom, or a themed party. Most of the time it moves gently, so it works as a long-running backdrop; the overload every half hour gives a scene a moment of tension. Teachers can pause it and use the panels as a starting point for talking about qubits, temperatures, and error rates. For OBS setup as a browser source, see the streaming toolkit.
A quantum computer is a machine that stores and processes information in physical systems governed by quantum mechanics. It is not a faster version of an ordinary computer. For most tasks it would be slower and far more expensive. For a few kinds of problem, though, algorithms are known that should let a large enough, accurate enough quantum computer do things no ordinary computer could do in any reasonable time. Building one that is large and accurate enough is still the hard part.
An ordinary bit is either 0 or 1. A qubit can be in a combination of 0 and 1 at the same time, called a superposition. The combination is described by two numbers called amplitudes, one for 0 and one for 1, and those numbers can be positive, negative, or complex. That is what makes the difference: amplitudes can add up or cancel out, like waves.
A popular description says a quantum computer "tries every answer at once". That is misleading. With many qubits the machine's state does involve an enormous number of amplitudes (two to the power of the number of qubits), but you cannot read them out. A quantum algorithm is a carefully designed sequence of operations, called gates, that makes the amplitudes of wrong answers cancel and those of right answers build up, so that a measurement at the end is likely to give something useful.
When qubits interact, they can become entangled: their joint state can no longer be described by giving each qubit its own state. Measure one, and the result is correlated with what you will find when you measure the other, in ways that no ordinary shared plan could reproduce. Entanglement is a resource that quantum algorithms rely on. The glowing links in the core stand for it, though real entanglement is not a visible beam and does not connect fixed pairs of neighbours.
Reading a qubit gives a plain 0 or 1, with probabilities set by its amplitudes, and the superposition is gone afterwards. So a single run of an algorithm gives one sample of possible answers. Real machines run the same circuit thousands of times, each run called a shot, and look at how often each result comes up. The result box on this screen quotes a number of shots for that reason. The amber and cyan pattern at measurement is the screen's way of showing ones and zeros.
Superposition and entanglement are fragile. Any stray interaction with the outside world (heat, vibration, electromagnetic noise, a passing cosmic ray) leaks information out of the qubits and destroys their quantum state. This is decoherence, and it is the main enemy of anyone building a quantum computer. Two standard measures of how long a qubit lasts are T1, how long it takes to lose energy and fall from 1 to 0, and T2, how long it keeps a well-defined phase. For superconducting qubits these are typically tens to hundreds of microseconds, so the 214 and 182 microseconds on the qubit map are in a realistic range. A real decoherence problem does not set off a containment alert. It shows up as more errors in the results.
The most common kind of qubit today is a tiny superconducting circuit on a chip, controlled with microwave pulses. For it to work, two things must be true. The metal must superconduct, which for materials such as aluminium needs temperatures around a degree above absolute zero or lower. And there must be so little heat around that random thermal energy almost never kicks a qubit out of its state, which needs temperatures far lower still.
That is the job of a dilution refrigerator. It cools in stages, each colder than the one above, much as the coolant panel shows: a plate at about 50 kelvin, one at about 4 kelvin, a still below 1 kelvin, a cold plate around 100 millikelvin, and finally the mixing chamber at around 10 to 20 millikelvin, where a mixture of helium-3 and helium-4 provides the last step of cooling. The processor is bolted to the coldest plate, with cables running down through every stage. Cooling one down or warming it up takes many hours to days. If a real fridge warmed up, the qubits would simply stop working; nothing would flash or vent.
Not every design needs this. Other approaches trap single charged atoms (ions) or neutral atoms in vacuum chambers and control them with lasers, or use particles of light (photons). Each has its own trade-offs in speed, accuracy, and how easily it can grow.
Physical qubits make mistakes far more often than ordinary transistors do, roughly one operation in a few hundred to a thousand on good current hardware, close to the one in a thousand shown on the screen. Quantum error correction spreads one reliable logical qubit across many physical ones and keeps measuring checks, called syndromes, that reveal errors without disturbing the stored information. A fast classical computer, the decoder, works out from the syndromes which correction to apply, about a million times a second.
The best-known scheme is the surface code, which arranges qubits on a grid. Its size is given by its distance, d: a bigger distance means more qubits and fewer logical errors, but only if the physical error rate is below a threshold, which for the surface code is roughly one error in a hundred operations. That is the dashed red line at 1 × 10−2 on the error rate panel. The screen's numbers do not quite add up, though: in the usual layout a distance-17 surface code needs about 2d² physical qubits, nearly 600, for a single logical qubit, which is more than the 408 this core claims to have.
| On this screen | In a real lab |
|---|---|
| A glowing ball of qubits in a hall | A chip a few centimetres across, hidden inside a refrigerator |
| Containment rings with magnetic coils | No rings; magnetic shielding keeps stray fields out rather than holding anything in |
| Core charge and overload | No charge builds up; a machine that misbehaves gives noisy results |
| Emergency vent and cold restart in 40 seconds | Warming up and cooling down take hours to days, and nothing is vented |
| A new answer about once a minute | Jobs wait in a queue, run thousands of shots, and are processed by ordinary computers afterwards |
The whole ten-hour cycle is worked out in advance from a fixed random seed: when each calculation cycle starts and how long it lasts, which job it runs and what invented result it gives, and when the decoherence alerts and overloads fall. From that schedule a single function works out the state of the core at any moment: its charge, how tight the rings are, how severe an alert is, and whether it is venting, dark, or rebooting. The qubits sit at fixed points on shells (spread evenly using a spiral pattern) and are rotated and projected from 3D onto the screen with a simple perspective camera. The glows are blended additively so that overlapping light builds up. Because every frame is a function of the time alone, the page can jump anywhere in the cycle at once, and the same moment always looks the same.
No. It is an animation drawn in your browser. The core, the calculations, the results, the temperatures, and every line in the log are invented. Nothing is computed on any quantum hardware, and the page connects to nothing.
No. This is a science-fiction picture. A real superconducting quantum processor is a chip a few centimetres across, mounted at the bottom of a dilution refrigerator: a tall cylinder that, opened up, shows a hanging stack of metal plates and cables. Nothing glows, nothing spins, and there is no core that can overload or vent. Other designs hold single ions or atoms in vacuum chambers with lasers, and look like optics benches.
Superconducting qubits only work when their circuits superconduct and when there is too little heat around to knock them out of their delicate states. Dilution refrigerators cool the chip to around ten to twenty thousandths of a degree above absolute zero, which is why the screen's coolant panel shows a mixing chamber at about 15 millikelvin.
Ten hours. The cycle holds 520 calculation cycles, 62 decoherence alerts, and 20 overloads. After ten hours it starts again from the beginning. The screen is calculated frame by frame rather than played from a recording, so it needs no download and never buffers.
Yes. Press N for the next event (a decoherence alert or an overload) and B for the previous one, or drag the seek bar. To open the page at a given moment, add ?t= to the address with the number of seconds, for example ?t=780 for the first overload.
Yes. It is free to use for videos, films, streams, music visuals, presentations, and events. Click FULLSCREEN, or add the page as an OBS browser source at 1920 × 1080 or 3840 × 2160; the scene is drawn at the full resolution of the source.
Yes, once every thirty minutes or so. At the end of each overload countdown the screen jumps to a pale white at about 60% strength and fades back over just over a second; then the core goes dark for half a minute. In the minute before, a red glow at the screen edges pulses, speeding up from about once every two seconds to about one and a half times a second, and the warning text blinks once a second. Each calculation cycle ends with a soft glow around the core that rises and falls over two or three seconds, and during decoherence alerts a few qubits flicker red. There is no strobing. If your system is set to reduce motion, the page opens paused. See accessibility and visual effects for more.
Once the page and its fonts have loaded, the simulation makes no further network requests, so it keeps running if the connection drops. The optional VIDEO view streams from YouTube and needs a connection.
For more control rooms, try the reactor control room, which has its own alarms and scram, or the tech dashboard for a busy wall of charts. The oscilloscope draws real waveforms from sound, and the hyperspace starfield is a calmer sci-fi backdrop. For machines on a bigger scale, see the AI data centre screen.