San Francisco, November 12, 2025 — IBM (NYSE: IBM) announced on Wednesday that it has developed a new experimental quantum computing chip called “Loon,” marking a major step toward achieving useful, error-corrected quantum computers by the end of the decade. The new chip showcases IBM’s continued progress on its quantum roadmap, which aims to move the technology from lab-scale experiments to real-world applications by 2029.
IBM’s Breakthrough: The Loon Quantum Chip
The Loon chip, unveiled by IBM Research, represents a key milestone in solving one of quantum computing’s biggest challenges — error correction. Quantum computers rely on qubits, which can represent both 0 and 1 simultaneously, enabling exponentially greater computing potential compared to classical systems. However, qubits are extremely fragile and prone to interference from their environment, leading to computational errors.
To address this, IBM began developing a hybrid error correction model in 2021, adapting an algorithm originally designed to improve cellphone signal processing and applying it to quantum data. This method uses both quantum and classical chips working together to detect and correct errors in real time.
“The Loon chip demonstrates that IBM’s quantum error correction architecture can be built into hardware — not just simulated,” said Jay Gambetta, Director of IBM Research and IBM Fellow. “It’s a crucial step toward realizing useful quantum machines within this decade.”
Why the Loon Chip Matters
The Loon chip isn’t yet ready for commercial use, but it’s proof of concept that IBM’s approach to integrating error-correcting qubit connections is viable. This integration makes quantum chips harder to manufacture, but far more powerful and stable once completed.
Mark Horvath, Vice President and Analyst at Gartner Research, called IBM’s achievement “a major leap forward.”
“It’s very, very clever,” Horvath told Reuters. “They’re not just theorizing anymore — they’re actually putting this architecture into physical chips. That’s super exciting.”
The Loon chip was fabricated at the Albany NanoTech Complex in New York, one of the world’s most advanced semiconductor research centers. The facility uses the same chipmaking tools and lithography systems found in top global foundries, allowing IBM to rapidly prototype next-generation quantum hardware.
The Path to Practical Quantum Computing
IBM’s latest announcement reinforces its belief that quantum advantage — the point at which quantum computers outperform classical supercomputers on specific tasks — is within reach.
In addition to Loon, IBM unveiled another new chip, “Nighthawk,” which will be available to researchers and developers later this year. According to IBM, Nighthawk could outperform classical computers on certain computational problems as early as 2026.
To ensure transparency and collaboration, IBM plans to open-source much of its quantum software stack. The company is partnering with startups, universities, and research labs to create a global quantum computing ecosystem where code and benchmarks can be shared publicly.
“We’re confident there’ll be many examples of quantum advantage,” Gambetta said. “But let’s move beyond headlines and papers — let’s create a community where researchers can submit their code, test results openly, and identify which problems quantum computers truly solve better.”
Competition in Quantum Computing
IBM is one of several technology giants racing to commercialize quantum computing. Alphabet’s Google, Amazon Web Services, and Intel are pursuing their own architectures and error-correction methods, each aiming to overcome the decoherence and scaling limitations that have kept quantum systems experimental for decades.
Analysts say IBM’s approach — integrating error-correcting interconnections directly into chip hardware — could provide an edge, allowing the company to scale quantum systems faster and more reliably than competitors relying solely on software-based correction.
Looking Toward 2029
IBM’s roadmap envisions a fully error-corrected quantum computer capable of handling commercially meaningful workloads by 2029. Such machines could revolutionize industries like drug discovery, materials science, cryptography, logistics, and artificial intelligence, where complex problems outstrip the capabilities of even today’s most powerful supercomputers.
While IBM did not specify when the Loon chip would be accessible to external partners, the company confirmed that select research groups will soon be able to test Nighthawk under the IBM Quantum Network, a cloud-based platform that already connects hundreds of organizations worldwide.
“This isn’t science fiction anymore,” Gambetta said. “We’re entering the phase where quantum computers start to become truly useful.”
Leave a Reply