On November 4, 2026, D-Wave Quantum announced a sharper version of its Stride hybrid solver, the cloud tool that mixes classical computing with quantum annealing to tackle large optimization problems. In key benchmark tests, the enhanced solver found better answers in 15 seconds than the previous version found in 120 seconds. Alongside the upgrade, D-Wave launched a three-course Applied Quantum Optimization learning path, aiming to turn a faster solver into a skill more developers can actually use.
What Stride does, and why optimization is hard
Optimization problems sound simple — find the cheapest route, the best schedule, the most efficient factory plan — but they explode in complexity as options multiply. A delivery network with thousands of stops, a factory with dozens of interdependent steps, or a workforce roster with shifting constraints can have far more possible answers than a classical computer can check one by one in a reasonable time. Stride approaches this differently. It combines classical computing resources with D-Wave’s quantum annealing hardware through the company’s Leap quantum cloud service, so users do not need to own a quantum computer to try it. According to D-Wave, the solver supports up to two million variables and constraints, covering workflows such as workforce scheduling, logistics routing, price optimization and production scheduling.
The headline benchmarks: 15 seconds versus 120 seconds
D-Wave’s announcement centres on three benchmark comparisons against the previous Stride solver. For travelling salesperson problems with up to 10,000 locations, solutions averaged within 0.02% of the best possible result in 15 seconds, and were on average more than 200 times closer to that best result than the previous solver at the same runtime. For problems with more than 10,000 locations, solutions averaged within 3.8% of the best possible result in 120 seconds, more than 500 times closer than the previous version at the same runtime. On simple assembly line balancing problems, the enhanced solver found better solutions in 15 seconds than the previous version found in 120 seconds. Those figures are vendor-reported benchmarks, and D-Wave itself notes that the improvement factors measure closeness to the optimal result, not raw processing speed — an important distinction when judging what the upgrade means in practice.
Where this kind of optimization already shows up
The value of a solver is proven in operations, not benchmarks, and D-Wave points to several real deployments. North Wales Police used Stride to optimize the placement of police vehicles for emergency response, with D-Wave reporting a nearly 50% reduction in average incident response time in that proof-of-technology project. BASF used the solver to optimize manufacturing workflows in a liquid-filling facility, cutting production scheduling time from 10 hours to seconds, according to D-Wave. A commercial application for Verge Ag calculated autonomous agriculture vehicle routes in one tenth of the previous time. These are company-reported results from specific projects, not guarantees for every workload, but they show the kinds of messy, constraint-heavy problems Stride is built for — the same class of problem that appears in logistics, manufacturing and emergency planning.
The new training path: Applied Quantum Optimization
A faster solver helps only if developers know how to frame a business problem for it, and that is the gap the new learning path targets. Applied Quantum Optimization is a three-course path offering hands-on instruction in identifying promising optimization use cases, formulating mathematical models, and building applications with Stride. It is designed for learners with varying levels of experience, and selected courses include access to the Leap quantum cloud service and office hours for practical learning. The courses are available now through Leap, with more detail on D-Wave’s training site. As with other recent efforts to put practical AI closer to real work — from on-device edge AI chips to new open-source models — the pattern here is clear: the tool matters less than whether working developers can use it on a real problem.
What to watch next, and the honest caveats
Three things will decide whether this upgrade matters beyond its launch-day benchmarks. First, independent results: vendor benchmarks compare the new solver with its own predecessor on chosen problems, not against every rival method, so outside tests on real customer workloads are the real verdict. Second, adoption: whether the training path brings in developers who were previously put off by quantum computing’s learning curve. Third, breadth: whether the gains seen on route planning and line balancing extend to the wider range of problems D-Wave says Stride supports. D-Wave senior vice president Alex Condello framed the goal in practical terms, saying developers need technology that finds better answers, faster, so they can tackle computationally complex problems at the speed their businesses demand. That is the right test — and it will be answered in production systems, not press releases.
Further reading
For context on how AI and optimization are moving into practical, everyday systems, see our report on Nordic’s nRF71 edge AI chips for battery-powered devices, Tencent’s Hy4 open-source AI model trained on defensive cybersecurity data, India’s homegrown EKA-1 and Dhīr AI models, and Astrolab’s LUMIS lunar power and communications node. For the source announcement, see D-Wave’s official announcement, HPCwire’s coverage of the Stride update, and D-Wave’s Leap hybrid solver page.
