Japan's room temperature quantum computer Shunkai goes live with 50 qubits

Started by WildManCena23, Sep 29, 2026, 04:31 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Topic: Japan's room temperature quantum computer Shunkai goes live with 50 qubits   Views(Read 97 times)

WildManCena23

Japan has switched on a full stack quantum computer called Shunkai that runs at room temperature. It was developed by the Institute for Molecular Science, part of the National Institutes of Natural Sciences, working with Hitachi and US company Infleqtion, which supplied the quantum processing unit. It currently runs with around 50 qubits. The plan is to grow that to about 500, with a target of a 10,000 qubit system by March 2031

Shunkai uses neutral atoms, where each qubit is a single atom held in place by laser light. Operations are carried out with microwave or laser pulses, and results are read by watching the light the atoms emit. The big selling point is that it does not need the giant cryogenic fridges that superconducting machines rely on to get close to absolute zero. That removes a lot of cost and complexity

Full stack means the system includes every layer needed to turn a user's input into a useful result, in the same way a laptop or phone hides all the complicated bits from you. Professor Kenji Ohmori said opening the machine to theory and software researchers, as well as corporate researchers, should create ripple effects towards practical applications. The main focus for now is developing and testing quantum error correction

IFLScience is honest that quantum computers still cannot beat top supercomputers at most tasks. Fifty qubits is modest compared with some of the machines out there, and scaling up remains the hard part. Still, neutral atoms have been one of the more promising routes to large systems, and having a working machine available to outside researchers is valuable

I like the room temperature angle, though it is worth remembering the atoms themselves are still cooled with lasers inside the machine. It is the surrounding hardware that stays warm. Does anyone think neutral atoms will end up winning the race, or will superconducting qubits stay ahead?


Coastal Otter

Error correction as the main focus is the right priority. Raw qubit counts mean little if errors pile up

A machine dedicated to testing correction schemes could help the whole field. Results should be published openly

Hollow85

Nice to see Japan back in the headlines for quantum. The US and China tend to dominate coverage, but Japan has been investing steadily for years. Riken already has machines running. Shunkai adds a different technology to their mix

Thor

Ten thousand qubits by 2031 is ambitious. Neutral atom arrays have scaled well so far, but controlling that many atoms precisely is another level

If they hit even half of that, it would be impressive. Japan does tend to deliver on long term science plans. Their track record on supercomputers is excellent
We can only hope this will be the last footwear to fall

Dean95

Infleqtion supplying the processor is interesting. A US company providing the core of Japan's flagship machine shows how international this field is. Hitachi brings the engineering and systems side. Each partner plays to its strengths. That kind of collaboration is probably how most of these machines will get built

George94

Neutral atoms have a real shot at winning. Qubits are identical by nature, which avoids the manufacturing variation that plagues superconducting chips. The weakness has been slower gate speeds

It will come down to which problem is easier to solve. My money is on atoms for large systems

Cougar

Worth noting that room temperature is slightly misleading. The atoms are laser cooled to extremely low temperatures inside a vacuum chamber. It is just that you avoid the huge cryogenic plumbing

Dataset Isaac

Full stack available to outside researchers is the most useful part. Access is what drives progress
My code works & I have no idea why

Griffin55

No dilution fridge is a big deal for cost. Those systems are expensive to buy and run, and they take up a lot of space. A machine that sits in a normal lab environment could be installed in far more places. That could widen access a lot
It's only trash talk... mostly