Google has introduced a new quantum computing chip called Willow that demonstrates an unprecedented milestone in quantum error correction.
It also completes a complex benchmark task in under five minutes, a calculation that would take the Frontier supercomputer about 10 septillion years.
The Frontier supercomputer by Hewlett Packard Enterprise (HPE), also known as OLCF-5, is the world’s first verified exascale supercomputer, capable of performing over one quintillion calculations per second (1.1 exaflops).
This achievement builds upon the company’s earlier quantum demonstrations, including the 2019 debut of the Google Sycamore chip. Willow’s success brings the industry closer to practical quantum devices capable of computations that classical hardware cannot feasibly undertake.
Scaling to 105 Qubits
The Willow chip, crafted at Google’s Santa Barbara quantum facility, incorporates 105 qubits, nearly doubling the 53 qubits of Google’s Sycamore 53-qubit CPU.
A qubit is the basic unit of information in quantum computing, similar to a regular computer bit, but it can exist in multiple quantum states simultaneously, allowing for more complex calculations.
Beyond simply adding more qubits, the researchers behind Willow proved that errors do not grow uncontrollably with scale.
Instead, by grouping physical qubits into logical units, Willow became more reliable as it grew larger. “In other words, we achieved an exponential reduction in the error rate,” said Dr. Hartmut Neven, founder and lead of Google Quantum AI.
“This historic accomplishment is known in the field as ‘below threshold’ — being able to drive errors down while scaling up the number of qubits.”
Researchers showed that creating logical qubits in progressively larger grids—from 3×3 to 5×5 to 7×7—reduced error rates exponentially. This confirms that once the quality of qubits crosses a specific line, scaling up does not magnify errors but instead curbs them.
Michael Newman, a research scientist at Google Quantum AI, referenced the origins of this idea: “Back in the 90s, it was first proposed that, if your qubits are just good enough, if they pass some kind of magical line in the sand called the quantum error correction threshold…” Now, Willow turns this theoretical threshold into an experimental fact.
Random circuit sampling (RCS) is the chosen test to verify beyond-classical performance. It’s a quantum computational task that involves sampling from the probability distribution of outcomes for randomly generated quantum circuits.
RCS is a leading candidate for demonstrating quantum supremacy, as it is designed to be prohibitively difficult for classical computers to simulate while being feasible for quantum computers to implement.
In 2019, Sycamore completed an RCS benchmark in just over three minutes, a task previously estimated to take a top-tier supercomputer thousands of years.
Now, Willow advances this benchmark dramatically. One RCS test that Willow performs in under five minutes would challenge Frontier for 10 septillion (10^25) years.
“If you can’t win on random circuit sampling you can’t win on any other algorithm either. The next task is to train this massive compute power to a task the mainstream would care for,” said Dr. Neven.
Longer Coherence and Practical Potential
Willow’s qubits maintain coherence—retaining their quantum states—around five times longer than earlier devices. These improved coherence times expand the window for complex computations and further enhance the chip’s stability.
While RCS proves that quantum machines can surpass classical limits, the next step involves applying their power to problems that matter to industry and research. Such real-world applications may include simulating complex molecules for drug discovery or modeling materials at the atomic level.
Toward Useful Quantum Computation
By revealing that quantum systems can run beyond-classical tasks without error rates spiraling out of control, Willow sets the stage for new practical uses. Google’s roadmap suggests that, over time, quantum computing may become as indispensable as today’s classical supercomputers.
Willow’s performance—achieving below-threshold error correction and far outpacing classical hardware—indicates that the long-envisioned era of useful quantum computation might finally be approaching.
