IBM announced earlier this year that its Condor quantum processor has crossed the 1,000-qubit threshold, a milestone that would have seemed extraordinary five years ago. Google's Willow chip has demonstrated quantum error correction that reduces error rates exponentially as qubits are added. And yet the honest answer to whether quantum computing has changed anything practical for businesses in 2026 is: not yet, but the trajectory is clearer than it has ever been.
The distinction between physical qubits and logical qubits matters enormously for understanding this gap. IBM's 1,000+ physical qubits are noisy — they decohere quickly and produce errors that require extensive classical correction. Estimates vary, but most researchers believe you need somewhere between 1,000 and 10,000 physical qubits per useful logical qubit, depending on the error correction scheme. That means IBM's current hardware might support a handful of logical qubits — not nearly enough to tackle problems that classical computers genuinely cannot solve.
Google's approach with Willow is different: the company is demonstrating that adding more physical qubits actually improves error correction rather than making it worse — a crucial prerequisite for scaling. The December 2024 paper in Nature showing Willow's error correction scaling was the most significant quantum result in years, not because the chip did anything immediately useful, but because it showed the physical principle works.
Where quantum computing is already relevant — if cautiously — is in quantum simulation for materials science and chemistry. Companies like Quantinuum (a Honeywell/Cambridge Quantum merger), IonQ, and IBM are running experiments for pharmaceutical companies and battery manufacturers to simulate molecular interactions that classical computers handle poorly. BASF and Johnson & Johnson have both confirmed quantum simulation partnerships. The outputs are not yet production-quality, but they are informing research directions.
The financial sector has explored quantum optimization for portfolio rebalancing and risk calculation, but classical algorithms running on accelerated hardware have proven competitive enough that quantum advantage in finance remains elusive at practical problem sizes.
For most businesses, quantum computing is a technology to monitor and prepare for rather than deploy. Building quantum literacy within engineering teams now — understanding what problems are genuinely quantum-amenable — is the practical advice that IBM and Google both give enterprise customers. The companies who will extract value from quantum computing first are those already experimenting with quantum algorithms on simulators, ready to migrate to hardware when the error rates come down to useful levels.
Timeline estimates have consistently been pushed back, but the physical progress in 2024-2026 has been real enough that most researchers now believe fault-tolerant quantum computing at meaningful scale is a question of when, not if — probably the early 2030s for the first commercially significant applications.