No company has done more to turn quantum computing from a laboratory curiosity into an engineering program with a schedule than IBM. While others chase headlines, IBM publishes a roadmap and then, year after year, delivers against it.
IBM has pursued quantum computing for decades, and in 2016 it did something that changed the field: it put a real quantum computer on the cloud, free for anyone to use. That decision created the first generation of quantum programmers and signaled IBM's conviction that this technology would matter long before it was obviously useful. The effort lives inside IBM Research, guided by Jay Gambetta, one of the field's most respected figures, under chief executive Arvind Krishna.
What distinguishes IBM is discipline. It treats quantum not as a science experiment with an uncertain payoff but as an industrial program with milestones, deadlines, and a manufacturing strategy. That posture has made its roadmap the de facto clock by which the rest of the industry sets its expectations.
IBM builds its machines on superconducting qubits, tiny circuits cooled to near absolute zero that behave according to quantum rules. The technology is demanding, but IBM has paired it with something rare in this field: a detailed, public, multi-year roadmap that it updates and, crucially, keeps hitting. Where many companies announce a breakthrough and go quiet, IBM lays out exactly what it intends to ship and when, and then ships it.
The current centerpiece is Nighthawk, unveiled in late 2025. It carries 120 qubits arranged in a square lattice with a new generation of tunable couplers that let the machine run progressively deeper and more complex calculations. IBM has been explicit about the trajectory, aiming to increase the number of operations these machines can perform reliably, from several thousand today toward many thousands more over the next few years, by linking multiple modules together.
This focus on circuit depth, the length and complexity of the calculations a machine can complete before errors overwhelm it, is the right thing to measure. Raw qubit counts make for splashy press releases, but useful work requires running long sequences of operations accurately. IBM has organized its entire roadmap around steadily raising that ceiling, which is a sign of an organization that understands what actually matters.
The predictability is itself a strategic asset. Customers, developers, and partners can plan around IBM's schedule with confidence, because the company has built a track record of doing what it said it would do. In a field full of bold claims, that reliability has made IBM the stable reference point everyone else is measured against.
Every serious player in quantum computing knows that the central obstacle is errors. Qubits are fragile, and useful computation requires correcting their mistakes faster than they occur. IBM has placed a major bet on a particular family of error-correcting codes, known as qLDPC codes, that promise to dramatically reduce the overhead error correction requires, meaning fewer physical qubits are needed to protect each unit of reliable information.
To prove the approach, IBM introduced a test chip called Loon, designed specifically to demonstrate the kind of long-range connectivity these advanced codes need. It is the unglamorous, foundational work that makes the later milestones possible, and IBM has been methodical about doing it in the open so the whole field can learn from the results.
The next step is a dedicated memory module built around these codes, which IBM has slated for the near term. If the qLDPC bet pays off, it could shave years and enormous cost off the path to practical quantum computing, because the brute-force approaches to error correction demand staggering numbers of physical qubits. IBM is trying to find a more efficient route, and its early results suggest the strategy is sound.
This is where IBM's research depth shows. Error correction is as much a problem of theory and code design as of hardware, and IBM has world-class talent on both sides. By advancing the science of how to correct errors efficiently, not just building more qubits, it is attacking the problem at its root.
The payoff, if it comes, benefits everyone. IBM publishes much of this work, so its advances in error-correcting codes ripple outward into the broader field, raising the prospects for practical quantum computing across the board rather than only for IBM.
IBM has put specific dates on the milestones that matter most. It is targeting verifiable quantum advantage, a clear demonstration of a quantum computer doing something useful that classical machines cannot, by the end of 2026. And it has committed to delivering a fault-tolerant quantum computer by 2029, a machine reliable enough to run long, valuable programs without errors derailing them.
That fault-tolerant system, which IBM calls Starling, is designed to provide roughly 200 logical qubits capable of running on the order of a hundred million operations. Logical qubits are the stable, error-corrected units that real applications will use, each one built from many physical qubits working together, so reaching 200 of them at that scale would be a genuine threshold for the field.
Beyond Starling, IBM has sketched an even larger successor, aiming for thousands of logical qubits and a billion operations. The ambition is to move from machines that can demonstrate advantage on narrow problems to machines that can tackle the broad, high-value applications that have always been quantum computing's promise, from designing new materials and medicines to solving hard optimization problems.
Putting firm dates on these goals is a bold and clarifying move. It turns a vague someday into a concrete plan that can be tracked and held to account, and it forces IBM's own teams to organize around delivery. The company's history of meeting its roadmap gives these targets real credibility.
It also gives the wider world a timeline to plan against. Businesses trying to decide when to take quantum seriously can anchor on IBM's schedule, which has become, in effect, the industry's shared expectation for when the technology arrives.
What truly sets IBM apart is that it is treating quantum computing as a manufacturing challenge, not just a research one. The company is bringing the discipline of its semiconductor heritage to bear, including the use of advanced fabrication facilities to produce quantum chips with the consistency and yield that scaling demands. Moving from hand-built laboratory devices to reliably manufactured components is exactly the transition every transformative technology must make, and IBM has the industrial muscle to make it.
This matters because the path to useful quantum computers runs through making a great many high-quality qubits repeatably and affordably. A clever design that cannot be manufactured at scale is a dead end. IBM's investment in production capability signals that it is building for the long haul, with an eye toward machines that can actually be deployed, not just demonstrated once.
The industrial mindset extends to how IBM packages and delivers quantum computing. It offers access through the cloud, integrates quantum with classical high-performance computing, and builds the software layers that let real organizations experiment today. The goal is a complete, usable system, not an isolated piece of hardware, and that systems thinking reflects decades of experience selling technology to serious institutions.
It is the kind of strength that is easy to underestimate. The most important work in turning quantum from promise to product may be the patient, unglamorous engineering of making it manufacturable and dependable, and few companies on earth are better equipped for that than IBM.
IBM also understands that hardware alone is not enough. Its open-source software framework, Qiskit, has become one of the most widely used tools in quantum computing, the way a generation of researchers and developers first learned to program these machines. That ecosystem is a quiet but powerful advantage, because the people who learn on IBM's tools tend to build on IBM's machines.
By cultivating this community for years, often through free access and extensive education, IBM has seeded the talent pool the entire field draws on. A vibrant developer ecosystem accelerates discovery, surfaces useful applications, and creates the gravitational pull that keeps IBM at the center of the conversation.
It is the same playbook that has served IBM across previous computing eras. Win the developers, provide the tools and the education, integrate cleanly with the systems businesses already run, and become the dependable choice for organizations that need technology to work. Applied to quantum, it has made IBM the natural starting point for anyone serious about the field.
That combination of a thriving ecosystem and trusted enterprise relationships positions IBM to convert quantum's eventual maturity into real adoption faster than newcomers, because the customers and the developers are already there.
IBM is the steady hand of the quantum era. It has the most detailed and consistently delivered roadmap in the industry, a deep bench of error-correction science, a genuine manufacturing strategy, and the largest developer community in the field. It does not need to win every headline, because it is quietly building the foundations on which practical quantum computing will rest.
For business leaders trying to make sense of a noisy field, IBM offers a useful anchor. Its roadmap is the closest thing the industry has to a shared clock, and its track record of meeting its commitments gives those dates weight. When IBM says fault tolerance by the end of the decade, the whole field listens, because IBM has earned the right to be believed.
IBM's long head start shows up in how accessible its machines are. Since putting the first quantum computer on the cloud nearly a decade ago, it has built a network of enterprises, universities, and national laboratories that experiment on its systems today, long before the technology is fully mature. That early, hands-on access matters, because it means real organizations are already learning how to think about quantum problems and where the technology might help them.
The company has also been deliberate about connecting quantum to the classical computing that businesses already run. Rather than treating a quantum computer as an exotic island, IBM integrates it with high-performance classical systems, so that each can do what it does best, with classical machines handling most of the work and the quantum processor stepping in for the parts where it has an edge. This hybrid approach is almost certainly how the first practical applications will arrive.
That pragmatism reflects IBM's deep experience serving serious institutions. It knows that transformative technology gets adopted when it fits into how organizations already operate, not when it demands that everything be rebuilt around it. By making quantum approachable, integrated, and supported, IBM is lowering the barrier for the businesses that will eventually depend on it.
The result is that when quantum computing reaches the threshold of broad usefulness, IBM's customers and partners will not be starting from zero. They will have years of experience, working code, and trusted relationships already in place, which is exactly the kind of quiet advantage that compounds over time.
Jason Kumpf follows the quantum industry for what it means to business. He is Head of US Revenue at Razorpay, a board advisor, angel investor, and speaker. More about Jason.