IBM Unveils Ambitious Roadmap for Fault-Tolerant Quantum Computer 'Starling' by 2029

IBM claims 'real world' edge in quantum computing race

Technology veteran IBM on Tuesday laid out a plan to have a "practical" quantum computer tackling big problems before the end of this decade. Current quantum computers are still experimental and face significant challenges, including high error rates. Companies like IBM, Google, and others are working to build more stable and scalable quantum systems. Real-world innovations that quantum computing has the potential to tackle include developing better fuels, materials, pharmaceuticals, or even new elements. However, delivering on that promise has always seemed some way off. IBM insists that it is now closer than ever and can map out the path toward having a quantum system that can fulfill actual use cases. This breakthrough comes thanks to a system being built in a data center in Poughkeepsie, New York, which IBM billed as a pioneering move toward making quantum an everyday work tool. The company's "Starling" quantum computer, targeted for completion in four years, will serve as the foundation for a next-generation machine ready by 2033 that will be dramatically more powerful in terms of the complexity of the problems it can tackle. "You've probably heard people talk about all these things, but we actually have plans for how to build it," IBM fellow and director of quantum systems Jerry Chow told AFP. The performance leap by Starling from today's systems will be on par with the jump from flip phones to smartphones, according to the company. Chow credited an innovative way to correct errors and ramp up fault tolerance with clearing the way for Starling. "People talk about error correction from the point of view of it being the Holy Grail," Chow said. "That really is the thing that is necessary to drive solutions at scale." While hailing IBM's "significant advancement," Gartner analyst Mark Horvath noted that challenges remain despite the milestone and said real-world applications for quantum systems remain "somewhat theoretical." Quantum AI? After several dashed predictions, quantum computing is accelerating rapidly. US tech giants, startups, banks, and pharmaceutical companies are pouring investments into this revolutionary technology. IBM announced plans in April to invest $150 billion in the United States over five years, including $30 billion earmarked for research and development to bolster manufacturing of mainframe and quantum computers. IBM has also outlined a "quantum-centric supercomputing" vision that involves using the new technology to augment traditional computer resources, letting each play to its strengths. If all goes to plan, quantum computing could start hitting its stride as generative artificial intelligence (AI) improves at breakneck speed, giving it incredible processing power. IBM claims its quantum services have generated sales of $1 billion, though current applications remain limited mainly to research. While classical computing relies on bits -- the smallest, most basic unit of digital information -- with just two states (0 or 1), quantum computing harnesses so-called qubits with infinite possible states. The "Bluejay" system, slated for completion four years after the "Starling" in 2033, will operate with 2,000 logical qubits and a billion "quantum gates" -- a technical reference to operations or parts of a problem it can handle, according to IBM. Each added qubit theoretically doubles computing capacity, allowing quantum computers to analyze countless possibilities simultaneously, solving in minutes what might take millions of years classically. "This is the most exciting thing that I can be spending my time on," said Chow, who has worked on quantum at IBM for 15 years. "We get to redefine what computation looks like for the future." IBM -- which has trailed Microsoft, Amazon, and Google for classic computing products in recent years -- claims the world's largest fleet of quantum computing systems. Google late last year showed off a new quantum computing chip that it said could bring practical quantum computing closer to reality. Its custom chip called "Willow" does in minutes what it would take leading supercomputers 10 septillion years to complete, according to the search engine giant. Tech giant Microsoft in February unveiled its own quantum computer chip, Majorana 1, saying it made the technology more reliable for real-world use.

IBM lays out roadmap for new mega quantum computer by 2029

IBM has a number of processors it hopes to release before Starling, which are set to culminate in its development. IBM aims to deliver the "world's first" large-scale, fault-tolerant quantum computer by 2029, and on Tuesday (10 June), it laid out its plans to achieve that goal. The goal, the Quantum Starling computer is expected to perform 20,000-times more operations than today's quantum machines, the company said. A quantum computer with hundreds or thousands of logical qubits could accelerate time and increase cost efficiency to boost drug development, materials discovery and many other fields. The Starling is meant to be a practical and scalable way for the next step in quantum computing. According to the company, the computational state of Starling would require the memory of more than a quindecillion (or, eight trillions multiplied together) of today's most powerful supercomputers. The new quantum computer will be built in a new IBM data centre in New York. IBM has a number of processors it hopes to release before Starling, which are set to culminate in the quantum computer's development. The Quantum Loon processor is expected this year, and is designed to test the architectural components for code that connects qubits over longer distances within the same chips. Next year, Quantum Kookaburra is set to mark the company's first modular processor. It will combine quantum memory with logic operations - the basic building block for scaling fault-tolerant systems beyond a single chip, according to the company. Starling will be able to access the computational power required for large-scale, intensive problems by running 100m quantum operations using 200 logical qubits. Plus, Starling will also be the foundation for IBM Quantum Blue Jay, which will be capable of executing 1bn quantum operations over 2,000 logical qubits, the company said. IBM already operates a global fleet of quantum computers. Earlier this year, the company reportedly told CNBC that it has "booked a billion dollars' worth of [cumulative] quantum business" over the years, with hopes to accelerate that momentum. "IBM is charting the next frontier in quantum computing," said Arvind Krishna, the company's chairperson and CEO. "Our expertise across mathematics, physics and engineering is paving the way for a large-scale, fault-tolerant quantum computer - one that will solve real-world challenges and unlock immense possibilities for business." Last year, IBM, along with NASA and Oak Ridge National Laboratory, shared an AI model designed to address challenges related to short-term weather and long-term climate projection in a flexible, scalable way. Don't miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic's digest of need-to-know sci-tech news.

IBM claims 'real world' edge in quantum computing race

Technology veteran IBM on Tuesday laid out a plan to have a "practical" quantum computer tackling big problems before the end of this decade. The company's "Starling" quantum computer, targeted for completion in four years, will serve as the foundation for a next-generation machine ready by 2033 that will be dramatically more powerful in terms of the complexity of the problems it can tackle.Technology veteran IBM on Tuesday laid out a plan to have a "practical" quantum computer tackling big problems before the end of this decade. Current quantum computers are still experimental and face significant challenges, including high error rates. Companies like IBM, Google, and others are working to build more stable and scalable quantum systems. Real-world innovations that quantum computing has the potential to tackle include developing better fuels, materials, pharmaceuticals, or even new elements. However, delivering on that promise has always seemed some way off. IBM insists that it is now closer than ever and can map out the path toward having a quantum system that can fulfill actual use cases. This breakthrough comes thanks to a system being built in a data center in Poughkeepsie, New York, which IBM billed as a pioneering move toward making quantum an everyday work tool. The company's "Starling" quantum computer, targeted for completion in four years, will serve as the foundation for a next-generation machine ready by 2033 that will be dramatically more powerful in terms of the complexity of the problems it can tackle. "You've probably heard people talk about all these things, but we actually have plans for how to build it," IBM fellow and director of quantum systems Jerry Chow told AFP. The performance leap by Starling from today's systems will be on par with the jump from flip phones to smartphones, according to the company. Chow credited an innovative way to correct errors and ramp up fault tolerance with clearing the way for Starling. "People talk about error correction from the point of view of it being the Holy Grail," Chow said. "That really is the thing that is necessary to drive solutions at scale." While hailing IBM's "significant advancement," Gartner analyst Mark Horvath noted that challenges remain despite the milestone and said real-world applications for quantum systems remain "somewhat theoretical." Quantum AI? After several dashed predictions, quantum computing is accelerating rapidly. US tech giants, startups, banks, and pharmaceutical companies are pouring investments into this revolutionary technology. IBM announced plans in April to invest $150 billion in the United States over five years, including $30 billion earmarked for research and development to bolster manufacturing of mainframe and quantum computers. IBM has also outlined a "quantum-centric supercomputing" vision that involves using the new technology to augment traditional computer resources, letting each play to its strengths. If all goes to plan, quantum computing could start hitting its stride as generative artificial intelligence (AI) improves at breakneck speed, giving it incredible processing power. "The potential is tremendous," Chow said. Ever more powerful Quantum computers already operate today. IBM claims its quantum services have generated sales of $1 billion, though current applications remain limited mainly to research. While classical computing relies on bits -- the smallest, most basic unit of digital information -- with just two states (0 or 1), quantum computing harnesses so-called qubits with infinite possible states. The "Bluejay" system, slated for completion four years after the "Starling" in 2033, will operate with 2,000 logical qubits and a billion "quantum gates" -- a technical reference to operations or parts of a problem it can handle, according to IBM. Each added qubit theoretically doubles computing capacity, allowing quantum computers to analyze countless possibilities simultaneously, solving in minutes what might take millions of years classically. "This is the most exciting thing that I can be spending my time on," said Chow, who has worked on quantum at IBM for 15 years. "We get to redefine what computation looks like for the future." IBM -- which has trailed Microsoft, Amazon, and Google for classic computing products in recent years -- claims the world's largest fleet of quantum computing systems. Google late last year showed off a new quantum computing chip that it said could bring practical quantum computing closer to reality. Its custom chip called "Willow" does in minutes what it would take leading supercomputers 10 septillion years to complete, according to the search engine giant. Tech giant Microsoft in February unveiled its own quantum computer chip, Majorana 1, saying it made the technology more reliable for real-world use.

Building Starling: IBM's fault-tolerant quantum computer coming in 2029

Starling's scale marks a milestone comparable to the advent of classical supercomputers Forgotten in the buzz of WWDC 2025, IBM's quantum computing announcement might not have been as flashy - but it was quite revealing, not unlike looking through clear liquid glass. IBM's announcement lays out a remarkably concrete roadmap to deliver Starling, the first large-scale, fault-tolerant quantum computer, by 2029. Housed in a new IBM Quantum Data Center in New York, Starling promises to execute some 20,000 times more quantum operations than any system in operation today. That's some marketing hyperbole, as representing its state would require memory equivalent to over a quindecillion or 10 of today's fastest classical supercomputers. Also read: Google Willow quantum chip explained: Faster than a supercomputer This isn't IBM's first quantum roadmap. But it is the first time the company has publicly laid out detailed processor milestones - Loon (2025), Kookaburra (2026), Cockatoo (2027), culminating in Starling (2029) - each testing critical elements of error-corrected, modular architectures based on quantum LDPC or qLDPC codes to slash qubit overhead by 90% relative to previous codes. In plain terms, IBM is tackling the quantum equivalent of the classical "vacuum tube to transistor" transition, which is nothing short of revolutionary, if you ask me. To appreciate Starling's significance, we need to rewind to the dawn of programmable computing. Time for a much-needed history lesson. In 1946, ENIAC, a 30-ton behemoth, calculated artillery tables at just 5,000 additions per second. A few years later, IBM's 650 drum-memory system, at roughly half a million dollars per unit, became the first mass-market computer, ushering in an era of batch processing for business and science. IBM's history is tightly woven into both classical and quantum threads. In the 1950s, IBM's transistor-based computers replaced vacuum tubes, notably with the IBM 1401, making data processing affordable for mid-sized businesses. A decade later, System/360 pioneered architecture-compatible families, defining the mainframe era. Also read: IBM reveals faster Heron R2 quantum computing chip: Why this matters Fast-forward to 2011, that's when IBM launched the first 5-qubit quantum processor at Yorktown Heights. Each subsequent leap - 16 qubits, 50 qubits - tested coherence times and basic algorithms but remained "noisy intermediate-scale quantum" (NISQ) devices, limited by error rates and decoherence. It wasn't until IBM's Nature-cover paper on quantum LDPC codes that a clear, efficient path to error-corrected logical qubits emerged . Now, with Starling and beyond, IBM is uniting decades of hardware, control electronics, and algorithm research into a practical vision: 100 million quantum operations on 200 logical qubits, followed by 1 billion ops on 2,000 logical qubits with Blue Jay in 2033. This roadmap is more than ambition - it's a stepwise de-risking of large-scale quantum computing, much like IBM's systematic moves in classical eras. Classical hardware (which is what all our modern PC, laptop, smartphones comprise) is remarkably reliable - bits stay intact and errors are rare. Quantum systems, by contrast, suffer from noise, where stray electromagnetic fields can flip qubit states in microseconds. Fault tolerance, then, is the foundational pillar of quantum computing's promise - where thousands of noisy physical qubits are bundled together into a single logical qubit, whose error rates shrink exponentially with cluster size. Also read: Amazon introduces Ocelot, its first quantum computing chip: All you need to know IBM's papers detail how qLDPC codes and real-time decoding with classical processors will keep these logical qubits coherent long enough to run meaningful algorithms. It's the quantum counterpart of error-correcting memory in classical RAM, but orders of magnitude more complex. Achieving this at scale - without insane overhead - was a longstanding hurdle. Starling promises to break that barrier. What's the need for quantum computing? Think of it this way, that classical computing gave us the internet, climate models, and AI that warns us of pandemics. But certain frontiers - like simulating complex molecules for drug discovery or solving massive optimization problems - remain tantalizingly out of reach. This is where quantum computers shine. IBM Quantum Starling stands as the next great expedition into the great unknown of computational limits. If delivered by 2029, it will let chemists model reactions at quantum fidelity, economists optimize vast portfolios, and cryptographers test new protocols. And just as the transistor's debut reshaped society, so might this first fault-tolerant quantum turn a scientific corner. Also read: Shaping the future of quantum computing: Intel's Anne Matsuura IBM's role in all of this is unique: pioneering transistor-era hardware, building mainframes that powered global commerce, and now forging quantum architectures. Few organizations possess the cross-disciplinary mastery to span vacuum tubes to qubits. Yet one can't help but be cynical about their ambitious timeline. Because classical progress took decades per leap. But IBM's quantum research team must compress that into a few short years lest other players - academia, startups, or global rivals - capture the lead. Yes, IBM isn't the only tech company fiddling with quantum computing breakthroughs - there's Google, Intel, Amazon, Microsoft and few others are also in the race. From ENIAC's vacuum tubes to the IBM System/360's transistor logic, from Google's Willow chip demonstration to today's IBM Starling blueprint, we're witnessing a continuum of human ingenuity. IBM's new roadmap doesn't just chart a technical path, but aspires to create a computing revolution. As we count down to 2029, the real question isn't just whether IBM can build Starling - it's whether we can harness its power responsibly to tackle the world's most pressing challenges.