VQAs for Today's Quantum Computing Hardware

Who has the best quantum processor today? Ask the physics community quietly and many will say: Quantinuum 𝗡𝗼𝘁 IBM Quantum. 𝗡𝗼𝘁 Google. 𝗡𝗼𝘁 IonQ. It’s easy to get caught up in roadmaps, qubit counts or quantum advantage headlines. But the real turning point for the field currently isn’t about scale. It's about fault tolerance - detecting and correcting quantum errors faster than they accumulate. Through that lens, Quantinuum’s H-Series trapped-ion system stands apart. Here’s why: • 𝗥𝗲𝗰𝗼𝗿𝗱-𝗛𝗶𝗴𝗵 𝗚𝗮𝘁𝗲 𝗙𝗶𝗱𝗲𝗹𝗶𝘁𝗶𝗲𝘀: The H-Series delivered a sustained >99.9% two-qubit gate fidelity and >99.99% single-qubit gate fidelity. This is the quality baseline to ensure any QEC code has a chance to work.    • 𝗟𝗼𝗴𝗶𝗰𝗮𝗹 𝗕𝗿𝗲𝗮𝗸-𝗘𝘃𝗲𝗻: They've repeatedly demonstrated logical qubits that are more reliable than the physical hardware they're built from—the first milestone for practical quantum computing.    • 𝗨𝗻𝗶𝘃𝗲𝗿𝘀𝗮𝗹 𝗚𝗮𝘁𝗲 𝗢𝗽𝗲𝗿𝗮𝘁𝗶𝗼𝗻𝘀: Achieved logical gate fidelity an order of magnitude better than physical fidelity on 𝗻𝗼𝗻-𝗖𝗹𝗶𝗳𝗳𝗼𝗿𝗱 𝗴𝗮𝘁𝗲𝘀, which are the hardest operations to perform fault-tolerantly and essential for universal quantum computing.    • 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗮𝗹 𝗔𝗱𝘃𝗮𝗻𝘁𝗮𝗴𝗲: All-to-All Connectivity. The Quantum Charge-Coupled Device (QCCD) system uses ion shuttling to provide full all-to-all qubit connectivity.    • 𝗧𝗵𝗲 𝗤𝗘𝗖 𝗧𝗲𝘀𝘁𝗯𝗲𝗱: This architecture allows them to deploy a diverse range of QEC codes (Steane, Carbon, Tesseract) and test protocols like Single-Shot QEC and Fault-Tolerant Teleportation. It is literally built to explore and accelerate the FTQC roadmap.

Logical qubits announcements are flying these days. Just last week: - IBM Quantum touted an experiment with "140 logical qubits": https://lnkd.in/eNYj7vZC - Quantinuum introduced their Helios computer with up to "94 logical qubits": https://lnkd.in/eEh2BuFQ So, is this it? Are we now in the logical qubit era? Well... we're entering it, but we're not quite there yet. If we go back to the basics, to be called a "quantum computer", a machine must meet DiVincenzo's criteria (https://lnkd.in/e5ETewiB): 1. A scalable physical system with well-characterized qubit 2. The ability to initialize the state of the qubits to a simple fiducial state 3. Long relevant Quantum coherence times 4. A "universal" set of quantum gates 5. A qubit-specific measurement capability Current logical qubit experiments demonstrate 1, 2 and 5. 3 is debatable, since "long" is ill-defined, but some experiments achieve better-than-physical performance, so let's consider we have it too. The real problem lies in 4: most of today's "logical qubits" are either quantum memories (i.e. they can preserve information but not manipulate it), or they feature some gates but not a universal set. This is not a small detail: if you cannot run a specific type of gates (non-Clifford ones) on your logical qubits, then you are left with circuits that can be classically simulated with just a polynomial overhead. In other words: without non-Clifford gates, quantum computing's exponential speedup is gone, and so is any hope of practical advantage. Now, I do agree with Jay Gambetta when he says "the only thing that matters is a quantum computer that runs a quantum circuit with a number of operations on a number of qubits" (in a discussion with Michaela Eichinger, PhD at https://lnkd.in/edJgWda4 ). But I would add: this needs to be true for any operation (or gate) one might want to run. Being able to run millions of Clifford gates is nice but insufficient. Of course, Jay and other industry leaders know this, and research teams are hard at work implementing logical non-Clifford gates. Some results have even already been published, for example here by Quantinuum: https://lnkd.in/esXuxhjj So, next time you see a large number of logical qubits being announced, don't stop at this figure: - Ask if there is a universal gate set - Ask how low error rates are - Oh, and ask how fast logical gates are With this additional information, you'll have a clearer understanding of the progress being made!

Nature Magazine just published a peer-reviewed study by SEEQC that represents a major breakthrough for quantum computing. It's now the journal's #1 most-read cover page article & in the top 2% of all research papers ever tracked. SEEQC has solved one of the foundational problems blocking large-scale quantum computers. Today's systems run control electronics at room temperature and pipe signals down individual wires into an ultra-cold cryostat — one wire per qubit. That architecture cannot scale. SEEQC demonstrated that digital control electronics can operate directly on-chip at the same millikelvin temperature as the qubits themselves, with one wire serving multiple qubits. When BlueYard first backed SEEQC, many saw this as the stuff of science fiction. Not anymore. Link in comments. John Levy Oleg Mukhanov Shu-Jen Han Matthew Hutchings BlueYard Capital Peter Read Hartmut Neven Gregory M. Bernstein Jason Palmer Daniel Franke Lorne Abony Ted Persson

PsiQuantum Achieves Breakthrough in Mass-Producing Light-Powered Quantum Chips American quantum computing startup PsiQuantum has announced a major breakthrough in manufacturing scalable photonic quantum chips, marking a significant step toward making practical quantum computing a reality. The company, which emerged from stealth mode in 2021, has been working on a light-powered (photonic) quantum computing approach, which was previously considered impractical due to hardware limitations. Why Photonic Quantum Computing? • Photonic quantum computers encode data in individual particles of light (photons), rather than in superconducting circuits like many other quantum systems. • This approach has key advantages: • Low noise compared to superconducting qubits. • High-speed operation due to the natural speed of light. • Seamless integration with fiber-optic networks, which could make quantum internet feasible. • However, the challenge has always been scaling up, as photons are difficult to control, detect, and stabilize in large-scale computations. PsiQuantum’s Breakthrough • In a paper published in Nature, the company unveiled a manufacturing process that enables large-scale production of photonic quantum chips. • The new hardware design solves key engineering problems, making it possible to reliably manipulate and measure photons at scale. • Unlike previous photonic quantum systems, which struggled with extreme hardware demands, PsiQuantum’s solution reduces errors and improves stability in complex computations. Implications for the Future of Quantum Computing • Scalability Achieved – This breakthrough could allow for mass production of quantum chips, removing a key bottleneck in commercial quantum computing development. • Quantum Networking Potential – With natural fiber-optic compatibility, photonic quantum computers could lead to highly secure quantum communications networks. • New Industrial Applications – The technology may soon be applied to optimization problems, cryptography, and materials science, revolutionizing industries that require complex simulations. The Bigger Picture PsiQuantum’s ability to mass-produce photonic quantum chips puts light-powered quantum computing in direct competition with other approaches, such as superconducting and trapped-ion quantum systems. If successful, it could make quantum computing more accessible, scalable, and commercially viable—a leap forward in the race to achieve practical quantum supremacy.

https://lnkd.in/g5uZ4z7G Azulene Labs was happy to contribute to this work led by Ryan LaRose at Michigan State University, alongside collaborators from NASA Ames Research Center, Brown University, Virginia Tech, The University of Texas at Austin, and Stanford University. This paper tackles a fundamental question: what will it actually cost to run quantum algorithms on problems that matter in biochemistry? We focused on metaphosphate hydrolysis — a simplified model of ATP hydrolysis, one of the most important reactions in all of biology, with direct implications for metabolism, cellular signaling, and cancer therapeutics. The study evaluates three quantum algorithms spanning three eras of quantum hardware: variational methods (ADAPT-VQE) for today's noisy devices, quantum Krylov for hopefully-near-future "MegaQuop" machines, and quantum phase estimation for fully fault-tolerant systems. A key finding is that variational approaches might already require few enough quantum resources to be feasible this decade, while the more powerful algorithms still face significant overhead, largely driven by the complexity of simulating time evolution for real molecular Hamiltonians. Importantly, classical pre-processing through Hamiltonian downfolding (DUCC) was essential for making the problem tractable across all three algorithms, reducing a 50-electron, 78-orbital system down to a manageable 44-qubit active space. This highlights something we think about a lot at Azulene Labs: the future of computational chemistry might not quantum OR classical — it might end up being the intelligent integration of both. The same principles of active space selection, correlation recovery, and efficient resource allocation that make quantum algorithms viable are exactly what drive advances in classical (non-quantum-computing) computational chemistry today. Ryan LaRose, Antonios Alvertis, Alan Bidart, Ben DalFavero, Sophia E. Economou, J. Wayne Mullinax, Mafalda Ramôa, Jeremiah Rowland, Brenda Rubenstein, Nicolas Sawaya, Prateek Vaish, Grant Rotskoff, Norm M. Tubman. Funded by the Wellcome Trust, among others. #ComputationalChemistry #DrugDiscovery #Biochemistry #QuantumChemistry #ATPHydrolysis #PharmaTech #HPC #ScientificComputing #DeepTech #BioTech

Today marks a historic milestone in quantum computing, as Microsoft and Quantinuum demonstrate the most reliable logical qubits on record. This breakthrough, with a logical error rate 800x better than the physical error rate, signifies a giant leap from the noisy intermediate-scale quantum (NISQ) level (Level 1 – Foundational) to Level 2 – Resilient quantum computing.   This progress is significant as logical qubits are only useful when they have a better error rate than physical qubits themselves. The number of physical qubits is a misleading metric; it’s not how many qubits, it’s how good they are and how resilient the quantum system is to errors.   Using the logical qubits we created, we were able to successfully perform multiple active syndrome extractions, which is when errors are diagnosed and corrected without destroying the logical qubits. Active syndrome extraction helps quantum computers stay reliable even when operations are imperfect.   With the promise of a hybrid supercomputing system powered by these reliable logical qubits, we’re paving the way for scientific and commercial breakthroughs that were once deemed impossible.  This achievement is a testament to the power of collaboration and the collective advancement of quantum hardware and software.   You can learn more from my post on the Official Microsoft Blog https://lnkd.in/gnDfcUV6 and the companion technical post on the Azure Quantum blog by Dennis Tom and Krysta Svore: https://lnkd.in/gMRVPG3s. #quantum #quantumcomputing #azurequantum

🚀 Microsoft Azure Quantum and Quantinuum announced today a new milestone in the race to practical quantum computing. Some time ago vendors changed direction from having more qubits to having better qubits. As with many things in live, useful quantum computers are a matter of quality rather than quantity when it comes to qubits. Having high quality qubits and the ability to manipulate them with low error rates is a requirement to create any useful quantum circuit, the equivalent to programs in classical computers. In April both partners announced having set up 4 logical qubits (=high quality qubits) from the combination of 30 physical qubits on a Quantinuum H-series machine. (https://lnkd.in/ePp5MC7t) Now they claim having achieved 12 logical qubits on a 56-qubit Quantinuum H2. All 12 logical qubits were entangled in a GHZ state with a circuit error rate of 0.0011. Scaling to error rates in the order of 10^-3 is great news. Practical quantum computers able to address the complex problems waiting for them will require much more than that, but we are on the way. https://lnkd.in/e6e6v8Tk

EeroQ researchers published new findings in Physical Review X about controlling individual electrons at temperatures above 1 Kelvin. Here's what they accomplished: Current quantum computers operate near 10 millikelvin. EeroQ demonstrated electron control at temperatures 100 times higher. Their approach uses electrons floating on superfluid helium, integrated with standard superconducting circuits. Why this matters for quantum computing: → Reduces extreme cooling requirements   → Uses existing quantum hardware infrastructure   → Creates a cleaner environment for qubit operations   → May help with scaling challenges Johannes Pollanen, EeroQ's cofounder, noted this "reduces a key barrier to scalable quantum computing." The company has been developing this electron-on-helium technology since 2017. The work validates theoretical predictions about using helium as a platform for quantum operations. The research addresses a practical problem: current quantum systems require expensive, complex cooling to near absolute zero temperatures. For those working in quantum computing: What cooling challenges do you face in your systems? ♻️ Repost to help people in your network. And follow me for more posts like this.

𝙄𝙨 𝙩𝙝𝙞𝙨 𝙩𝙝𝙚 ‘𝙏𝙧𝙖𝙣𝙨𝙞𝙨𝙩𝙤𝙧 𝙈𝙤𝙢𝙚𝙣𝙩’ 𝙛𝙤𝙧 𝙩𝙝𝙚 𝙌𝙪𝙖𝙣𝙩𝙪𝙢 𝘼𝙜𝙚? Quantum computing has been full of promise, but real-world impact has always felt decades away. That changes now. Microsoft just unveiled Majorana-1, the world’s first quantum chip powered by a Topological Core architecture, a revolutionary breakthrough that could scale quantum computing from today’s small, fragile systems to a fault-tolerant, million-qubit supercomputer. 𝗪𝗵𝘆 𝗶𝘀 𝘁𝗵𝗶𝘀 𝗯𝗶𝗴? Think of how semiconductors enabled modern computers, smartphones, and AI. Microsoft’s topoconductor, a new quantum material that stabilizes qubits by controlling elusive Majorana particles, could do the same for quantum computing. 🔹 Qubits that are stable by design – drastically reducing errors, the #1 bottleneck in quantum computing 🔹 Scalability to a million qubits – the threshold required to solve real-world industrial and scientific problems 🔹 Pioneering a new computing paradigm – making the impossible possible in materials science, healthcare, and AI 𝗤𝘂𝗮𝗻𝘁𝘂𝗺 + 𝗔𝗜 = 𝗧𝗵𝗲 𝗡𝗲𝘅𝘁 𝗗𝗶𝘀𝗿𝘂𝗽𝘁𝗶𝗼𝗻 A million-qubit system could reinvent computing as we know it: ✔ AI models trained on quantum-optimized architectures ✔ New materials engineered to self-repair ✔ Drugs designed at the molecular level with unprecedented precision ✔ Climate solutions that break down microplastics and capture carbon Quantum isn’t just an experiment anymore. It’s getting ready for scale. 𝐌𝐢𝐜𝐫𝐨𝐬𝐨𝐟𝐭 𝐢𝐬 𝐨𝐧𝐞 𝐨𝐟 𝐭𝐰𝐨 𝐜𝐨𝐦𝐩𝐚𝐧𝐢𝐞𝐬 𝐚𝐝𝐯𝐚𝐧𝐜𝐢𝐧𝐠 𝐭𝐨 𝐭𝐡𝐞 𝐟𝐢𝐧𝐚𝐥 𝐩𝐡𝐚𝐬𝐞 𝐨𝐟 𝐃𝐀𝐑𝐏𝐀’𝐬 𝐔𝐒2𝐐𝐂 𝐩𝐫𝐨𝐠𝐫𝐚𝐦, aimed at delivering the first utility-scale fault-tolerant quantum computer, a machine whose computational value exceeds its costs. The future of computing isn’t just classical or AI. It’s quantum, and it’s coming faster than we thought. Read more: 🔗 https://lnkd.in/e5ztYJdi -- ♻️ Repost this to help others in your network. And follow Anurag(Anu) Karupartifor more!