Google's Advances in Fault-Tolerant Quantum Computing

Yesterday, Google announced it had achieved something called a “verifiable quantum advantage”. The announcement might sound like marketing mush but it’s not. It represents one of the most interesting inflection points in the story of computing since the transistor. For decades, the dream of quantum computing has dangled like science fiction: machines that use the strange rules of quantum mechanics to solve problems that would take supercomputers millennia. In 2019, Google claimed quantum supremacy - meaning their quantum computer solved a problem no classical computer could feasibly do in a reasonable timeframe. But that problem was a glorified dice roll: random number sampling. A proof of principle, not of purpose Their latest claim - quantum advantage - goes further. It says a quantum machine has outperformed the best classical algorithms on a task that’s scientifically meaningful. In their experiment, Google’s Willow processor, a 105-qubit superconducting chip, ran an algorithm called Quantum Echoes to model how information spreads and decoheres - essentially, how order unravels into chaos inside quantum systems. That’s the kind of math that underpins chemistry, materials science, and condensed-matter physics. Willow completed the task 13,000x faster than the world’s best supercomputers, while remaining verifiable - that is, its output could be independently checked. In other words, the machine wasn’t playing a party trick anymore; it was doing science. Every era of computing begins with a strange, narrow demo that later looks obvious in hindsight. ➰ The Wright brothers’ first flight lasted 12 seconds - not exactly air travel. ➰ The first transistor amplified a single signal - not exactly an iPhone. ➰ The first webpage looked like a grocery list - not exactly the internet. Google’s quantum milestone feels the same. A narrow, technical victory that, decades later, we’ll point to and say: that’s when the impossible started to feel inevitable. Of course, the hype shouldn’t outrun the hardware. Quantum systems face 3 towering challenges: ▪️ Error correction: Qubits are noisy - one stray photon can flip a bit of reality. ▪️ Scalability: Doubling qubits isn’t like doubling transistors; coherence decays exponentially. ▪️ Integration: Quantum systems must coexist with classical infrastructure - data movement, cooling, algorithms, verification. For now, the near horizon is hybrid quantum-classical computing, where quantum processors handle intractable subproblems inside classical workflows. For the past 80 years, computing has been about logic - zeros and ones manipulating symbols. Quantum computing is about reality itself: entanglement, superposition, uncertainty. It represents a paradigm where the map is the territory - where we use the universe’s own rules to understand the universe. In that sense, the shift from quantum supremacy to advantage mirrors the shift from theory to instrument - from “it works” to “it works for us.”

For those tracking progress in Quantum… As my colleague Hartmut Neven has predicted, real-world applications possible only on quantum computers are much closer than people think – as near as five years, even though fully error corrected quantum computers may be further away.  Recently, my colleagues on our Quantum AI team at Google Research took another important step on that path with a new set of results we published last week in Nature that share a promising new approach to applications on today’s quantum computers. Our analog-digital quantum simulator using super-conducting qubits shows performance beyond the reach of classical simulations in cross-entropy benchmarking experiments. Simulations with the level of experimental fidelity in this simulator would require more than a million years on a Frontier supercomputer. The simulator brings together digital’s flexibility and control with the analog’s speed – and provides a path towards applications that cannot be accomplished on a classical computer. Along the way, my colleagues also made a scientific discovery – they observed the breakdown of a well-known prediction in non-equilibrium physics, the Kibble-Zurek mechanism - an important result in our understanding of magnetism, and also useful in various kinds of quantum simulations. Congratulations to Trond Andersen, Nikita Astrakhantsev, and the rest of the team on this exciting step – much more to come! You can read the Nature paper here: https://lnkd.in/gg2En5qe 

One of the biggest challenges in quantum computing has always been error correction. Unlike classical computers, where errors are rare and manageable, quantum systems are incredibly sensitive. Even the tiniest disturbance can disrupt a calculation. For decades, scientists feared that error correction might require so much effort that it would outweigh the benefit of the computation itself—a roadblock for practical quantum computing. This week, Google announced a major breakthrough with its new #Willow chip, showing that error correction doesn’t have to diverge. They demonstrated that their system can perform calculations with 105 qubits, while simultaneously using error correction to manage and stabilize the system. For the first time, the overhead required for error correction scales in a manageable way as the system grows. Here’s why it’s game-changing: • 70 physical qubits are allocated to error correction for every logical qubit in the system, making the calculations reliable without overwhelming the computational capacity. • It proves quantum systems can become reliable at scale, bringing us closer to real-world applications like drug discovery, clean energy breakthroughs, and revolutionary materials design. • The Willow chip has already shown it can handle complex calculations that today’s fastest supercomputers couldn’t solve in the entire lifetime of the universe. Even Elon Musk couldn’t help but react, commenting “Wow” on X when the news dropped. This marks a turning point for quantum computing—it’s no longer just theoretical. The pieces are falling into place for a future where these machines solve humanity’s toughest problems. #AI #quantum

Very exciting new work out from Google, showcasing huge advances in the foundations of Quantum Error Correction. The protocol works as anticipated, showing that the process of encoding one qubit worth of information into many physical devices can deliver improvements to hardware over many repeated trials. Congrats to the team at Google for their continued impressive contributions to quantum information science! The next step is making this amazing scientific accomplishment practically useful for running #quantumalgorithms on #quantumcomputers. This passage from the Google Blog stood out to us at Q-CTRL: "At current physical error rates, we might need more than a thousand physical qubits per surface code grid to realize relatively modest encoded error rates of 10-6. Furthermore, all of this was accomplished on a 105-qubit processor; can we achieve the same performance on a 1,000-qubit processor?" https://lnkd.in/gwG3_VyS That is, the resource intensity of #QEC means that we have to do everything we can to reduce errors first before starting to implement QEC if we're seeking practical performance gains. Given current systems only have ~100 qubits, there's a lot of work to be done in order to make QEC relevant in a practical sense. This gets at the heart of our interests - making hardware perform better with infrastructure software is the key to making QEC practically useful! To learn more about the next steps on the journey to making QEC useful, have a look at our primer on the topic! https://lnkd.in/gbYrKCVj

Is Google’s Quantum Error Correction Already Outdated? Google Quantum AI’s December breakthrough in error correction marked a significant milestone toward building practical quantum computers. The team demonstrated that quantum computations could be corrected by grouping qubits, mitigating errors that typically multiply as quantum systems grow. However, their approach is already facing competition, raising questions about its future relevance. The Promise and Challenge of Quantum Computing Quantum computers are heralded for their potential to tackle complex problems in fields like materials science, chemistry, and logistics—tasks beyond the reach of classical systems. However, quantum computers are inherently error-prone due to their reliance on delicate quantum states that are easily disrupted by environmental noise. Error rates increase as quantum systems scale, making error correction a critical hurdle. Google’s Breakthrough Using their Willow quantum processor, Google demonstrated a method to reduce computational errors by grouping qubits into logical units. These grouped qubits could correct errors within themselves, marking progress toward “fault-tolerant” quantum computing. This achievement supports the notion that large-scale quantum computers capable of solving real-world problems are within reach. Rising Competition and Alternative Approaches Despite Google’s progress, competing methods are gaining attention and could potentially surpass its approach: 1. Superconducting vs. Alternative Platforms: Google’s quantum system uses superconducting qubits, but rival platforms like trapped ions, neutral atoms, and photonic qubits may offer better error tolerance or scalability. Companies such as IonQ and Atom Computing are advancing these alternative technologies. 2. Optimized Error Correction Algorithms: Some researchers are exploring algorithms that promise greater efficiency or compatibility with different quantum architectures, potentially making Google’s qubit-grouping method less relevant over time. 3. Post-Moore’s Law Techniques: Emerging methods that integrate classical and quantum error correction systems—such as hybrid computing models—may offer a more adaptable framework for quantum error correction. Future Implications While Google’s breakthrough represents a step forward, the rapid pace of quantum computing research means that today’s innovations could quickly become outdated. Competing technologies and alternative error correction methods may chart more efficient paths to the same goal: a scalable, error-free quantum computer. Google’s achievement, however, underscores a critical point—quantum error correction is not just a theoretical possibility but a tangible, progressing reality. The race to make quantum computing reliable and practical remains highly dynamic, with many players vying to define the future of the field.