World's first silicon-based quantum computer plugs into regular socket

[Chip (OP)]

https://www.livescience.com/technology/computing/worlds-first-silicon-based-quantum-computer-is-small-enough-to-plug-into-a-regular-power-socket?utm_term=032043BB-1CB4-4440-A845-2FF7DCCBD37B&lrh=1e7f7a9239bb44f191dc979b8fe5e634e587dfe020b84a653d2040468a8b342b&utm_campaign=368B3745-DDE0-4A69-A2E8-62503D85375D&utm_medium=email&utm_content=15C9EE73-953C-45FF-A0BD-246D75546B5A&utm_source=SmartBrief

World's first silicon-based quantum computer is small enough to plug into a regular power socket

7 May 2025

An Irish startup has created the world's first silicon-based quantum computer — it can still integrate seamlessly with classical computing in data centers.


cRw9szsLkmYENL9tLmSnfZ-970-80.jpg.webp (87.43 kB . 970x546 - viewed 19 times)
This new quantum computer uses silicon-based spin qubits as opposed to superconducting qubits or trapped-ion qubits deployed in most other machines. (Image credit: Fergal Phillips)

A startup has launched the first quantum device in the world that blends the potential of quantum computing with the convenience and integration of traditional high-performance computing (HPC).

Equal1 representatives unveiled Bell-1 on March 16 — a new six-qubit machine that can fit seamlessly into existing HPC environments like data centers, company representatives said in a statement.

The machine tips the scale at slightly more than 440 pounds (200 kilograms) but it's rack-mountable — meaning it can be mounted onto a physical rack in a data center — and it's roughly the same size as existing graphics processing unit (GPU) servers.


https://cdn.jwplayer.com/previews/UKzuAweh

Unlike other quantum computers, the Bell-1 doesn't require specialized infrastructure to deploy, and it doesn't need additional equipment to be cooled to near-absolute zero.

That's because it boasts its own self-contained, closed-cycle cryo cooling unit, which enables the system to operate at a remarkable 0.3 kelvin, or minus 459.13 degrees Fahrenheit (minus 272.85 degrees Celsius).

The machine makes use of the latest semiconductor fabrication techniques as well as purified silicon, which allows for a high level of control and long coherence times (a qubit's ability to exist in multiple states simultaneously, which is crucial for quantum algorithms and computations).

Rack-mountable quantum computing:

The Bell-1's qubits are silicon-based, meaning they’re smaller than conventional qubits, and the chip at the heart of the machine incorporates quantum processor units (QPUs) with Arm CPUs — traditional processors known for their small size and efficiency — and neural processing units (NPUs) — specialized processors for accelerating machine learning and artificial intelligence (AI).

Incorporating all these elements onto a single chip eliminates the complex orchestration that would otherwise be necessary between classical and quantum computing elements.

As long as you've got the space in a rack, all it requires is a standard electrical outlet: plug it in and it's ready to work, Equal1 representatives said.


DeUKZjB37vsVNss2JNnwdZ-970-80.jpg.webp (42.6 kB . 970x546 - viewed 13 times)
(Image credit: Fergal Phillips)

The company's chip, called the UnityQ 6-Qubit Quantum Processing System, utilizes spin qubits, in contrast to many quantum computing platforms that rely on either trapped-ion or superconducting qubits.

Silicon-based spin qubits are compact, leading to potentially higher qubit density, and could leverage existing semiconductor fabrication techniques, meaning more scalability.

The chip fitted into the Bell-1 also incorporates error correction, control and readout, while taking advantage of existing semiconductor infrastructure for reliability and scalability.

The Bell-1 builds on advances first published by the company in December 2024, which established new peak performance marks for silicon qubit arrays as well as quantum controller chips.

These included the world’s highest recorded single-qubit and two-qubit gate fidelity (meaning fewer errors) and gate speed (meaning faster operations).

The platform also utilizes a specialized, AI-powered error correction system developed in partnership with Arm.