Quantum computer with 98 atoms achieves record accuracy, pushing closer to practical use
A new trapped-ion quantum computer built by Quantinuum has set a benchmark for accuracy and scale, achieving error rates low enough to perform meaningful long calculations.
In a laboratory in Broomfield, Colorado, 98 atoms suspended in mid-air by electric fields and cooled near absolute zero form Helios, a new quantum computer that has achieved a significant milestone in the race toward practical quantum computing.
Built by the British-American company Quantinuum, Helios represents a shift in how the field measures progress. Rather than simply counting qubits—the quantum bits that store information—the breakthrough lies in the accuracy of those qubits, according to reporting from The Conversation citing a paper published in Nature.
The scale-and-accuracy challenge
Quantum computers operate on principles that have no parallel in classical computing. Their qubits can exist in quantum states that behave fundamentally differently from the ones and zeroes of conventional digital technology, enabling certain calculations to potentially outperform even the largest supercomputers. Possible applications range from designing new materials and optimizing complex systems to simulating chemistry and developing new cryptographic methods.
But qubits are fragile. Temperature fluctuations, imperfect controls, unwanted environmental interactions, and the movement of information around the device all introduce errors. This fragility means that having more qubits alone does not guarantee a better quantum computer.
Helios has 98 qubits, surpassing the previous largest system, Model H2, which had 56 qubits. But the real significance lies elsewhere.
Error rates that matter
According to the Nature paper, Helios achieves an average error rate of about 2.5 per 100,000 for single-qubit gates—the basic building blocks of quantum circuits. For two-qubit gates, which are more complex and essential for useful computation, the average error rate is about 7.9 per 10,000, matching the best demonstrations in the field at around 5 per 10,000 errors.
"The race in quantum computing is not only about having more qubits. It is about having more good qubits, controlled accurately enough to perform long and meaningful calculations."
These error rates are low enough to permit the kind of extended, complex calculations that have long been the bottleneck in quantum computing. As The Conversation framed it, judging quantum progress by qubit count alone is like judging a race by the number of runners at the starting line—what matters is how many finish, and in what condition.
What happens next
Helios demonstrates that trapped-ion quantum computers—which suspend atoms in electromagnetic fields rather than using other technologies like superconducting qubits—can deliver both scale and precision simultaneously. The achievement underscores a shift in the field: quantum computing announcements are increasingly focused on error correction and qubit quality rather than raw numbers.
Whether this represents the breakthrough the field has long promised remains uncertain. But the combination of 98 functional qubits operating at record accuracy suggests the practical applications of quantum computing may be moving from theory closer to reality.