Quantum Error Correction Set for 2028, Earlier Than Expected
▼ Summary
– Quantum computing announcements have been made this summer, including a promise of useful error-corrected quantum computing by 2028 and updates on trapped ion processors.
– Useful quantum computing requires error correction using logical qubits, which need thousands of high-quality hardware qubits.
– Current qubit technologies offer either high quality or large numbers, but not both, requiring incremental progress over several years.
– The 2028 promise is notably earlier than the common industry estimate of five to ten years away for useful quantum computers.
– Claims of quantum supremacy have been reduced due to improvements in traditional algorithms.
Quantum computing announcements often surge near year-end as companies race to demonstrate they are meeting their milestones. This summer, however, has delivered a wave of notable news, ranging from steady incremental advances to bold, headline-grabbing forecasts. Following our earlier roundup this month, Ars is highlighting several of the most impactful developments.
Among the standout claims is a promise to deliver useful, error-corrected quantum computing as early as 2028. Additional updates include details on an upgraded trapped ion processor, and a notable instance where claims of quantum supremacy have been tempered by progress in classical algorithms.
2028 is remarkably soon for such a breakthrough. Industry experts generally agree that practical quantum computers remain roughly five to ten years away. While a handful of useful algorithms might run on today’s error-prone machines, nearly all of the compelling applications for quantum computing require error correction. This involves linking a small set of hardware qubits into a single logical qubit, which stores information redundantly and uses neighboring qubits to detect and correct errors.
To perform meaningful computations, a system needs a healthy number of logical qubits. For instance, modeling simple chemical reactions might demand about 100 logical qubits, while complex tasks like breaking encryption could require tens of thousands. (Of course, the definition of “useful” depends heavily on who is asking.) This translates to a need for thousands of high-quality hardware qubits at a minimum.
Currently, qubit technologies tend to offer either high quality or large quantities, but not both. Existing roadmaps show a path forward, but they rely on years of steady, incremental progress. That is why the industry’s consensus estimate has remained at five to ten years. A target of 2028 would represent a dramatic acceleration of that timeline.
(Source: Ars Technica)
