background of a technical drawing of an ion trap and vacuum chamberNEWIonQ Introduces Roadmap, Algorithmic Qubit Metric

A true quantum leap

Introducing the first commercial trapped ion quantum computer. By manipulating individual atoms, it has the potential to one day solve problems beyond the capabilities of even the largest supercomputers.

Get Started

The World’s Most Powerful Quantum Computer

Our next-generation system features 32 qubits, minuscule gate errors, and an expected quantum volume greater than four million. Next year, it arrives on the cloud to accelerate research into solving humanity's hardest problems in chemistry, medicine, finance, logistics, and more.
Learn more about our latest breakthrough →

abstract icon representing power


Unparalleled gate depth

At the microscopic scales of quantum mechanics, processes are noisy. Near-term quantum computers are not inherently error-corrected; instead, noisiness limits the number of operations they can perform before their results are no longer meaningful. Our fully-connected, low-error qubits allow us to perform longer, more powerful calculations than any other system on the market.
abstract icon representing connectivity


All-to-all, random-access connectivity

Qubits are the basic building blocks of quantum computers. They are most powerful when they are directly connected, or entangled. (For an introduction to qubits and entanglement, see Ars Technica’s excellent article.) Because our qubits are fully connected in a static array, we can peform a logic gate between any pair of qubits, at any time. No other commercial quantum system has this capability.
abstract  icon representing connectivity


Ultra-low gate error
Fewer errors allow for longer and more useful calculations. For example, if our quantum logic gates had 20% error, we could apply 3 gates before our circuit output became too noisy to be useful. With 1% error, more than sixty gates are possible. At 0.1% error, this number jumps to over 700.

Atoms make better quantum computers

technical illustration of a vacuum system

Flexible Ion Trap Technology

In an ultra-high vacuum chamber,The atmospheric pressure we experience on the surface of earth is about 760 torr. By comparison, the vacuum of space in near-earth orbit is ~10-6 torr. For quantum computing, we aim for 10-11 torr or better, to maximize isolation of our qubits from the environment.we dynamically deploy and trap atomic qubits on a silicon chip using electromagnetic fields. This allows our quantum cores to adjust their configuration in software, and scale to handle potentially hundreds of qubits without new hardware.

technical illustration of a linear ion trap

Atomic Precision

Unlike other systems that try to simulate the behavior of atomic particles, IonQ quantum cores compute using many identical ytterbium atoms.Ytterbium is a silvery rare earth metal first discovered in the late 19th century. As of 2018, ytterbium is used to make some of the most accurate atomic clocks in the world. For example, PTB built a single-ion clock one hundred times more accurate than the best cesium clocks. To create our qubits, we ionize ytterbium 171, removing one of its outermost electrons. Want to learn more about how it all works? Drop us a line or consider studying at UMD’s Joint Quantum Institute!As in ytterbium atomic clocks, isolating individual atoms reduces error and improves stability.

technical illustration of laser optics

Laser Control

Precise lasers store information on our atomic qubits, perform logical operations, and connect them together in a quantum process called entanglement. With no fixed wires, the IonQ system can connect any two qubits with a single laser operation, increasing accuracy. By way of example, with full connectivity, when testing the Bernstein-Vazirani algorithm with a 10-qubit oracle, in the worst case we need simply 10 two-qubit gates. In a grid topology, with each qubit having 4 neighbors, you’d need 16 two-qubit gates. With a ring topology, you'd need 19 two-qubit gates. Since each additional gate introduces noise and thereby error, the fewer gates required, the higher quality the calculation.