A Bold Timeline for Fault-Tolerant Quantum Computing
Quantum computing firm QuEra says it plans to deliver a fault-tolerant quantum computer by 2028 and make it available through the cloud via a collaboration with Amazon Web Services. If achieved, the milestone would mark one of the most important transitions in the field: from experimental quantum hardware that remains too error-prone for many useful workloads to a system that can detect and correct its own mistakes at meaningful scale.
The company’s planned machine, called Libra, would be built from extremely cold neutral atoms controlled with lasers. According to the supplied source text, Libra is intended to contain between 10,000 and 15,000 qubits, partitioned into 256 logical qubits. QuEra projects that each logical qubit would commit an error only once in a million operations, enough to support a “megaquop,” or one million operations.
Why Error Correction Is the Real Barrier
Quantum computing’s most persistent technical problem is not the absence of qubits. It is the fragility of those qubits. Today’s systems are error-prone enough that even when they demonstrate impressive capabilities, the number of reliable steps they can execute remains limited. That sharply restricts their usefulness in chemistry, materials science and drug discovery, the very domains most often cited as future beneficiaries of quantum advantage.
A fault-tolerant machine would change that by continuously catching and correcting errors before they overwhelm a computation. That is why QuEra’s claim matters. The company is not just promising a bigger machine. It is promising a qualitatively different kind of machine, one that could move the industry closer to dependable application-scale work.
The Gap Between Today and 2028
The challenge is large. The source text notes that the biggest neutral-atom array currently contains 6,100 qubits, but those have not yet been used for computations. Meanwhile, the largest number of error-corrected logical qubits achieved so far is 48. QuEra’s roadmap would therefore require progress on several fronts at once: more physical qubits, stronger error correction, system integration and cloud-ready operational reliability.
That is why outside experts in the article treat the goal as plausible in direction but demanding in execution. Jonathan King of Atom Computing said a fully functional computing system, as opposed to a laboratory demonstration, will require many scientific and engineering advances to come together. In other words, the issue is not whether the underlying concept makes sense. It is whether all the enabling layers can mature quickly enough.
An Industry Race With Tight Timelines
QuEra’s 2028 target is also more aggressive than some rivals. The source text notes that IBM expects to offer fault-tolerant quantum computers starting in 2029. That means QuEra is trying to move ahead not only by solving the technical problem, but by doing so on a compressed schedule in one of the most capital- and expertise-intensive areas of modern computing.
The company’s choice of neutral atoms is central here. Neutral-atom systems offer a promising path because lasers can manipulate many atoms at once, potentially supporting large-scale arrays. But promise at the hardware architecture level does not eliminate the engineering complexity of turning those arrays into stable, repeatable and commercially useful machines.
What Success Would Mean
If Libra arrives on schedule and performs as advertised, it would be a landmark for the entire sector. A cloud-accessible, fault-tolerant system would give researchers and companies a platform for testing whether long-anticipated quantum use cases can finally move beyond theory. It would also put pressure on competing architectures and vendors to prove that their own roadmaps can deliver comparable reliability.
For now, though, QuEra’s announcement is best read as an ambitious engineering commitment rather than a settled fact. The field has a long history of optimism meeting hardware reality. Still, the race has clearly advanced. Companies are no longer talking only about whether fault tolerance matters. They are putting dates on when they think they can deliver it. That alone signals a new phase for quantum computing, even if 2028 turns out to be harder than planned.
This article is based on reporting by New Scientist. Read the original article.
Originally published on newscientist.com