QuEra says it is stepping off the “sell today, improve later” treadmill. In its roadmap, the quantum computing startup is making a strategic decision not to sell NISQ systems anymore, with a useful machine targeted for sometime in 2028 and a more powerful system arriving just one year later.

The key punchline is the size of the leap. QuEra’s current hardware is built around about 260 qubits, described as relatively error-prone. The roadmap points to a future system that, compared to what is on the market today, promises eye-popping performance: over 10,000 individual qubits with error rates low enough that the system could support hundreds of error-corrected logical qubits. Bridging that gap is not a minor upgrade. It is a fundamentally different regime, and QuEra is using timing and product decisions to manage the risk.

To understand why this matters, it helps to look at the Amazon angle that set expectations. Ars previously covered Amazon’s announcement that it would be hosting a useful quantum computer from its partner QuEra as soon as sometime in 2028. The implied bet is clear for decision-makers: if a cloud provider is willing to align its timeline with a quantum milestone, that milestone had better be credible. And credible, in quantum, is mostly about whether the system can move from manipulating qubits that are noisy and short-lived into doing computations that remain meaningful after errors accumulate.

QuEra’s starting point underscores the difficulty. The company has previously made two systems available, each with similar capabilities: about 250 hardware qubits and an appreciable error rate. Those systems are good enough to test error correction codes, but not sufficient for using logical qubits in applications. That distinction is crucial. Testing error correction codes is not the same as running error-corrected logical qubits in ways that can outperform classical approaches or deliver commercially relevant workloads. So when QuEra says it will not sell NISQ systems anymore, it is not just changing a sales motion. It is implicitly acknowledging that selling “in-between” hardware has limits, especially if buyers are looking for business outcomes.

QuEra’s Yuval Borger, speaking to Ars, framed the decision as strategic. The move also clarifies how QuEra intends to bridge the gap from its current ~260 qubit, relatively error-prone hardware to a machine that could support hundreds of error-corrected logical qubits. According to Ars, the roadmap announcement came last Wednesday and highlighted the gap rather than shrinking it. There will be no new hardware releases between now and the useful machine. In other words, the company is not promising incremental product steps that fans, researchers, or enterprise customers can immediately adopt. Instead, it is focusing on a near-term blackout of hardware launches to avoid underselling the eventual target or overpromising on the path there.

For executives, this creates both a constraint and a signal. The constraint is time. If you are evaluating quantum vendors, the absence of new releases means you cannot easily buy your way into improved performance on a rolling schedule. The signal is that QuEra believes the trajectory matters more than short-term shipments. That matters for cloud partners, system integrators, and investors trying to evaluate whether quantum is entering an era of “usable” capability rather than perpetual demonstrations.

There is also a market and capital implication hiding in plain sight. When a company stops selling NISQ systems, it reduces a revenue channel that may have supported deployments, pilots, or early research access. To do that, QuEra is effectively banking on a later payoff, namely the 2028 timeline that Amazon is aligning to, and a more powerful system one year after that. In boardroom terms, it is a trade: fewer near-term commercial artifacts in exchange for focus and credibility around a specific milestone. For investors and strategic partners, that can be a positive if the roadmap execution is strong, or a red flag if the lack of interim releases reflects engineering uncertainty rather than disciplined prioritization.

Finally, the competitive subtext is hard to ignore. Ars describes QuEra as promising it will leapfrog everybody, but the roadmap itself accentuates the gap: there is no hardware series of steps in between. In a field where many teams are racing on qubit counts and error rates, a decision to stop selling NISQ systems is an aggressive stance. It tells peers that the company is not competing on “how many qubits we can ship this quarter.” It is competing on how quickly it can turn qubits and error rates into error-corrected logical qubits that can support meaningful applications.

For decision-makers watching this, the stakes are practical. If your organization is planning for quantum talent, partnerships, or workloads that might benefit from error-corrected logical qubits, QuEra’s roadmap suggests the window is narrower and more milestone-driven than many NISQ-era conversations implied. And if you are allocating capital to quantum infrastructure, the message is that timelines and product strategy are now part of the technology itself, not just marketing around it.