Ģ����Ƶ

Today, we celebrate the role that quantum science and technology plays in our everyday lives. At Ģ����Ƶ, we are on a mission to use quantum technologies to positively impact the world. With our H-Series quantum computers and world-leading software solutions, we will continue to work with the global community of researchers, engineers, and scientists to push boundaries and solve industry’s most complex computational problems.

Interested in learning more about Ģ����Ƶ and our quantum computing technologies? Meet our team of quantum experts at these events: 

• From April 13^th - 15^th, our team of evangelists, business leaders and scientists will be participating in , Yale’s inaugural quantum computing hackathon, and hosting an associated workshop for students and professors. Participants will have the opportunity to design new quantum algorithms and use quantum algorithms to solve unique problems.

• On April 15^th, Prof. Bob Coecke will be in Washington, DC to discuss quantum inclusivity through “Quantum in Pictures”, a groundbreaking curriculum created to democratize quantum skills. Join Prof. Coecke at Elevate Quantum’s “” event, established to brief Members of Congress, staffers and agency leaders on the state of quantum, workforce development and the need to build an inclusive quantum community.

• On April 17^th, Anand Shah will deliver a talk on “” at the C4IR KSA event in Saudi Arabia. Anand will share insights on the quantum computing ecosystem, Ģ����Ƶ’s full-stack quantum computing technology and our work with partners and collaborators to develop industry-relevant applications.

Hear from our experts on the value of provable quantum randomness for cybersecurity, in partnership with Thales

The world is preparing for the biggest cryptographic migration in history. Organizations must update their tools and policies to provide robust crypto discovery, post-quantum algorithms, and sources of provable quantum randomness. Starting early is critical to ensure migration is complete before the quantum threat materializes.

Join Thales, IBM and Ģ����Ƶ as we delve into practical guidance for a post-quantum transformation on May 8, at 10:30am in the North Hall Briefing Center at Quantum Leap: Insights and Approaches for a Post-Quantum World. Speakers include Todd Moore, Global Head of Data Security Products at Thales, Antti Ropponen, Executive Partner and Global Data & Application Security Services Leader, and Duncan Jones, Head of Cybersecurity at Ģ����Ƶ.

In addition to speaking alongside Thales and IBM at the RSA Conference, Ģ����Ƶ is sponsoring Thales’ PQC Palooza on May 8 from 4:00 – 7:00pm at the Hyatt Regency Hotel. This is an exclusive post-quantum cryptography (PQC) event with industry experts sharing their approaches to preparing for the post-quantum era. The keynote address will be delivered by industry luminary Dr. Taher Elgamal, one of the forefathers of SSL, who will provide his unique perspective on cryptography and the upcoming PQC transformation. Additionally, there will be a panel discussion and an opportunity to talk with the Ģ����Ƶ team.

Visit our team at Booth 5280

5280 is the elevation (in feet) of Denver, where our commercial quantum computer sits. It’s also the number of our booth at the RSA Conference, where you can speak with our cyber experts. Look for our booth #5280 in the North Hall.

We look forward to seeing you in San Francisco!

By Ilyas Khan, Chief Product Officer and Jenni Strabley, Senior Director Offering Management

Ģ����Ƶ and Microsoft have announced a vital breakthrough in quantum computing that as “a major achievement for the entire quantum ecosystem.”

By combining Microsoft’s innovative qubit-virtualization system with the unique architectural features and fidelity of Ģ����Ƶ’s System Model H2 quantum computer, our teams have demonstrated the most reliable logical qubits on record with logical circuit error rates 800 times lower than the corresponding physical circuit error rates. 

This achievement is not just monumental for Ģ����Ƶ and Microsoft, but it is a major advancement for the entire quantum ecosystem. It is a crucial milestone on the path to building a hybrid supercomputing system that can truly transform research and innovation across many industries for decades to come. It also further bolsters H2’s title as the highest performing quantum computer in the world.

Entering a new era of quantum computing

Historically, there have been widely held assumptions about the physical qubits needed for large scale fault-tolerant quantum computing and the timeline to quantum computers delivering real-world value. It was previously thought that an achievement like this one was still years away from realization – but together, Ģ����Ƶ and Microsoft proved that fault-tolerant quantum computing is in fact a reality.

In enabling today’s announcement, Ģ����Ƶ’s System Model H2 becomes the first quantum computer to advance to Microsoft’s Level 2 – Resilient phase of quantum computing – an incredible milestone. Until now, no other computer had been capable of producing reliable logical qubits. 

Using Microsoft’s qubit-virtualization system, our teams used reliable logical qubits to perform 14,000 individual instances of a quantum circuit with no errors, an overall result that is unprecedented. Microsoft also demonstrated multiple rounds of active syndrome extraction – an essential error correction capability for measuring and detecting the occurrence of errors without destroying the quantum information encoded in the logical qubit. 

As we prepare to bring today’s logical quantum computing breakthrough to commercial users, there is palpable anticipation about what this new era means for our partners, customers, and the global quantum computing ecosystem that has grown up around our hardware, middleware, and software. 

Collaborating to reach a new era

To understand this achievement, it is helpful to shed some light on the joint work that went into it. Our breakthrough would not have been possible without the close collaboration of the two exceptional teams at Ģ����Ƶ and Microsoft over many years.

Building on a relationship that stretches back five years, we collaborated with Microsoft Azure Quantum at a very deep level to best execute their innovative qubit-virtualization system, including error diagnostics and correction. The Microsoft team was able to optimize their error correction innovation, reducing an original estimate of 300 required physical qubits 10-fold, to create four logical qubits with only 30 physical qubits, bringing it into scope for the 32-qubit H2 quantum computer.

This massive compression of the code and efficient virtualization challenges a consensus view about the resources needed to do fault-tolerant quantum computing, where it has been routinely stated that a logical qubit will require hundreds, even thousands of physical qubits. Through our collaboration, Microsoft’s far more efficient encoding was made possible by architectural features unique to the System Model H2, including our market-leading 99.8% two-qubit gate fidelity, 32 fully-connected qubits, and compatibility with Quantum Intermediate Representation (QIR).

Thanks to this powerful combination of collaboration, engineering excellence, and resource efficiency, quantum computing has taken a major step into a new era, introducing reliable logical qubits which will soon be available to industrial and research users.

Understanding today’s error correction breakthrough

It is widely recognized that for a quantum computer to be useful, it must be able to compute correctly even when errors (or faults) occur – this is what scientists and engineers describe as �ڲ��ܱ���-�ٴDZ�������Գ���.��

In classical computing, fault-tolerance is well-understood and we have come to take it for granted. We always assume that our computers will be reliable and fault-free. Multiple advances over the course of decades have led to this state of affairs, including hardware that is incredibly robust and error rates that are very low, and classical error correction schemes that are based on the ability to copy information across multiple bits, to create redundancy. 

Getting to the same point in quantum computing is more challenging, although the solution to this problem has been known for some time. Qubits are incredibly delicate since one must control the precise quantum states of single atoms, which are prone to errors. Additionally, we must abide by a fundamental law of quantum physics known as the no cloning theorem, which says that you can’t just copy qubits – meaning some of the techniques used in classical error correction are unavailable in quantum machines. 

The solution involves entangling groups of physical qubits (thereby creating a logical qubit), storing the relevant quantum information in the entangled state, and, via some complex functions, performing computations with error correction. This process is all done with the sole purpose of creating logical qubit errors lower than the errors at the physical level.

However, implementing quantum error correction requires a significant number of qubit operations. Unless the underlying physical fidelity is good enough, implementing a quantum error correcting code will add more noise to your circuit than it takes away. No matter how clever you are in implementing a code, if your physical fidelity is poor, the error correcting code will only introduce more noise. But, once your physical fidelity is good enough (aka when the physical error rate is “below threshold”), then you will see the error correcting code start to actually help: producing logical errors below the physical errors. 

Ģ����Ƶ’s fault-tolerance roadmap

Today’s results are an exciting marker on the path to fault-tolerant quantum computing. The focus must and will now shift from quantum computing companies simply stating the number of qubits they have to explaining their connectivity, the underlying quality of the qubits with reference to gate fidelities, and their approach to fault-tolerance.

Our H-Series hardware roadmap has not only focused on scaling qubits, but also developing useable quantum computers that are part of a vertically integrated stack. Our work across the full stack includes major advances at every level, for instance just last month we proved that our qubits could scale when we announced solutions to the wiring problem and the sorting problem. By maintaining higher qubit counts and world class fidelity, our customers and partners are able to advance further and faster in fields such as material science, drug discovery, AI and finance.

In 2025, we will introduce a new H-Series quantum computer, Helios, that takes the very best the H-Series has to offer, improving both physical qubit count and physical fidelity. This will take us and our users below threshold for a wider set of error correcting codes and make that device capable of supporting at least 10 highly reliable logical qubits. 

A path to real-world impact

As we build upon today’s milestone and lead the field on the path to fault-tolerance, we are committed to continuing to make significant strides in the research that enables the rapid advance of our technologies. We were the real-time quantum error correction (meaning a fully-fault tolerant QEC protocol), a result that meant we were the first to show: repeated real-time error correction, the ability to perform quantum "loops" (repeat-until-success protocols), and real-time decoding to determine the corrections during the computation. We were the first to create non-Abelian topological quantum matter and braid its anyons, leading to .

The native flexibility of our QCCD architecture has allowed us to efficiently investigate a large variety of fault-tolerant methods, and our best-in-class fidelity means we expect to lead the way in achieving reduced error rates with additional error correcting codes – and supporting our partners to do the same. We are already working on making reliable quantum computing a commercial reality so that our customers and partners can unlock the enormous real-world economic value that is waiting to be unleashed by the development of these systems. 

In the short term – with a hybrid supercomputer powered by a hundred reliable logical qubits, we believe that organizations will be able to start to see scientific advantages and will be able to accelerate valuable progress toward some of the most important problems that mankind faces such as modelling the materials used in batteries and hydrogen fuel cells or accelerating the development of meaning-aware AI language models. Over the long-term, if we are able to scale closer to ~1,000 reliable logical qubits, we will be able to unlock the commercial advantages that can ultimately transform the commercial world. 

Ģ����Ƶ customers have always been able to operate the most cutting-edge quantum computing, and we look forward to seeing how they, and our own world-leading teams, drive ahead developing new solutions based on the state-of-the-art tools we continue to put into their hands. We were the early leaders in quantum computing and now we are thrilled to be positioned at the forefront of fault-tolerant quantum computing. We are excited to see what today’s milestone unlocks for our customers in the days ahead.

For more information

• Please register for Microsoft’s upcoming with Ģ����Ƶ
• Visit Ģ����Ƶ InQuanto to explore this state-of-the-art chemistry platform that will soon offer reliable logical qubits

Quantum computing promises to revolutionize everything from machine learning to drug design – if we can build a computer with enough qubits (and fault-tolerance, which is for a different blog post). The issue of scaling is arguably one of the hardest problems in the field at large: how can we get more qubits, and critically, how can we make all those qubits work the way we need them to? 

A key issue in scaling is called the “wiring problem”. In general, one needs to send control signals to each qubit to perform the necessary operations required for a computation. All extant quantum computers have a hefty number of control signals being sent individually to each qubit. If nothing changes, then as one scales up the number of qubits they would also need to scale up the number of control signals in tandem. This isn’t just impractical (and prohibitively expensive), it also becomes quickly impossible - one can’t physically wire that many signals into a single chip, no matter how delicate their wiring is. The wiring problem is a general problem that all quantum computing companies face, and each architecture will need to find its own solution.

Another key issue in scaling is the “sorting problem” - essentially, you want to be able to move your qubits around so that they can “talk” to each other. While not strictly necessary (for example, superconducting architectures can’t do this), it allows for a much more flexible and robust design – it is the ability to move our qubits around that gives us “all-to-all connectivity”, which bestows a number of advantages such as access to ultra-efficient high density error correcting codes, low-error transversal gates, algorithms for simulating complex problems in physics and chemistry, and more. 

Ģ����Ƶ just put a huge dent in the scaling problem with their using a clever approach to minimize the number of signals needed to control the qubits, in a way that doesn’t scale prohibitively with the number of qubits. Specifically, the scheme uses a fixed number of (expensive) analog signals, independent of the number of qubits, plus a single digital input per qubit. Together, this is the minimum amount of information needed for complete motional control. All of this was done with a new trap chip arranged in a 2D grid, uniquely designed to have a perfect balance between the symmetry required to make a uniform trap with the capacity to break the symmetry in a way that gives “direction” (eg left vs right), all while allowing for efficient sorting compared to keeping qubits in a line or a loop. Taken together, this approach solves both the wiring and sorting problems – a remarkable achievement.

Stop-motion ion transport video showing loading an 8-site 2D grid trap with co-wiring and the swap-or-stay primitive operation. Single Yb ions are loaded off screen to the left, and are then transported into the grid top left site and shifted into place with the swap-or-stay primitive until the grid is fully populated. The stop-motion video was collected by segmenting the primitive operation and pausing mid-operation such that Yb fluorescence could be detected with a CMOS camera exposure.

Stop-motion ion transport video showing a chosen sorting operation implemented on an 8-site 2D grid trap with the swap-or-stay primitive. The sort is implemented by discrete choices of swaps or stays between neighboring sites. The numbers shown (indicated by dashed circles) at the beginning and end of the video show the initial and final location of the ions after the sort, e.g. the ion that starts at the top left site ends at the bottom right site. The stop-motion video was collected by segmenting the primitive operation and pausing mid-operation such that Yb fluorescence could be detected with a CMOS camera exposure.

“We are the first company that has designed a trap that can be run with a reasonable number of signals within a framework for a scalable architecture,” said Curtis Volin, Principal R&D Engineer and Scientist.

The team used this new approach to demonstrate qubit transport and sorting with impressive results; demonstrating a swap rate of 2.5 kHz and very low heating. The low heating highlights the quality of the control system, while the swap rate demonstrates the importance of a 2D grid layout – it is much quicker to rearrange qubits on a grid vs qubits in a line or loop. On top of all that, this demonstration was done on three completely separate systems, proving it is not just “hero data” that worked one time on one system, but is instead a reproducible, commercial-quality result. Further underscoring the reproducibility, the data was taken with both Strontium/Barium pairs and Ytterbium/Barium pairs. 

This demonstration is a powerful example of Ģ����Ƶ’s commitment and capacity for the full design process from conception to delivery: our team designed a brand-new trap chip that has never been seen before, under strict engineering constraints, successfully fabricated that chip with exquisite quality, then finally demonstrated excellent experimental results on the new system. 

“It’s a heck of a demonstration,” quipped Ian Hoffman, a Lead Physicist at Ģ����Ƶ.

The is the world’s largest physics conference enhancing education and collaboration in a variety of scientific research areas. As the interest in and potential of quantum technology increases, so does the number of conference sessions about the topic.

This year, the Ģ����Ƶ team will be participating in many of the APS March Meeting sessions to discuss the latest advancements in quantum technology. Find us throughout the week at the below sessions and visit us at Booth 605 in the .

Join these sessions to discover how Ģ����Ƶ is advancing quantum computing

(A51) Applications on Noisy Ģ����Ƶ I
‍Speaker: Kentaro Yamamoto, Senior Research Scientist
‍Date: Monday, March 4th
Time: 9:24 a.m. - 9:36 a.m. CST

(A40) Probing Structure and Dynamics with XUV and X-Ray Light: Ultrafast Studies of Photocatalysis and Water Radiolysis
Speaker: Evgeny Plekhanov, Quantum Physics Research Scientist
‍Date: Monday, March 4^th
‍Time: 10:00 a.m. - 10:12 a.m. CST

(G30) Commercial Applications of Quantum Computing
Speaker: Nathan Burdick, R&D Manager
Date: Tuesday, March 5^th
Time: 12:42 p.m. – 1:18 p.m. CST

(G56) Scaling Trapped Ion Quantum Computers
Speaker: Ian Hoffman, Lead Physicist
Date: Tuesday, March 5^th
Time: 12:06 p.m. – 12:42 p.m. CST

(K49) Algorithms and Implementations on Near-Term Quantum Computers
Speaker: Matthew DeCross, Advanced Physicist
Date: Tuesday, March 5^th
Time: 3:36 p.m. – 3:48 p.m. CST

(Q51) Co-evolution of Quantum and Classical Algorithms
‍Speaker: Cristina Cirstoiu, Quantum Research Scientist
‍Date: Wednesday, March 6thTime: 3:00 p.m. - 3:36 p.m. CST

(Q49) Quantum Algorithms for Many-Body Systems
‍Speaker: Maria Tudorovskaya, Research Scientist
‍Date: Wednesday, March 6^th
‍Time: 5:12 p.m. - 5:24 p.pm. CST

(Q14) Quantum Many-Body Scars and Red Phenomena
‍Speaker: Michael Schecter, Senior Advanced Physicist
Date: Wednesday. March 6th
Time: 5:24 p.m. – 5:36 p.m. CST

(S53) Trapped Ion Qubits
Speaker: Robert Tyler Sutherland, Lead Physicist
Date: Thursday, March 7^th
Time: 8:00 a.m. – 8:36 a.m. CST

(S53) Trapped Ion Qubits
Speaker: Andrew Schaffer, Advanced Physicist
Date: Thursday, March 7^th
Time: 8:48 a.m. – 9:00 a.m. CST

(S53) Trapped Ion Qubits
Speaker: Adam Reed, Senior Advanced Physicist
Date: Thursday, March 7^th
Time: 9:12 a.m. – 9:24 a.m. CST

(S53) Trapped Ion Qubits
Speaker: Lucas Sletten, Advanced Physicist
Date: Thursday, March 7^th
Time: 10:12 a.m. – 10:24 a.m. CST

(S51) Quantum Error Correction Code Performance and Implementation II
Speaker: Ben Criger, Senior Research Scientist
Date: Thursday, March 7^th
Time: 10:24 a.m. – 10:36 a.m. CST

(T50) Applications on Noisy Ģ����Ƶ II
Speaker: Eli Chertkov, Advanced Physicist
Date: Thursday, March 7^th
Time: 12:30 p.m. – 12:42 p.m. CST

(T50) Applications on Noisy Ģ����Ƶ II
Speaker: David Hayes, Senior R&D Manager
Date: Thursday, March 7^th
Time: 12:42 p.m. – 1:18 p.m. CST

Interested in a career at Ģ����Ƶ? Meet our team at the Job Expo

Always on the leading edge of their fields, our hardware, software, sales, business, and operations teams are focused on personal, business, and technological growth. Curious, driven, and talented, our people are what makes Ģ����Ƶ tick. Every one of us is motivated to deliver on our mission to accelerate quantum computing. We are looking for team members with the same ambitions to join us!

Visit us at the to talk about positions at Ģ����Ƶ.

How do machines “learn”? 

While recent years have seen incredible advancements in Artificial Intelligence (AI), no-one really knows how these ‘first-gen’ systems actually work. New work at Ģ����Ƶ is helping to develop different frameworks for AI that we can understand - making it interpretable and accountable and therefore far more fit for purpose. 

The current fascination with AI systems built around generative Large Language Models (LLMs) is entirely understandable, but lost amid the noise and excitement is the simple fact that AI tech in its current form is basically a “black box” that we can’t look into or examine in any meaningful manner. This is because when computer scientists were starting to figure out how to make machines ‘human like’ and ‘think’, they turned to our best model for a thinking machine, the human brain. The human brain essentially consists of neural networks, and so computer scientists developed artificial neural networks. 

However, just as we don’t fully understand how human intelligence works, it’s also true that we don’t really understand how current artificial intelligence works – neural networks are notoriously difficult to interpret and understand. This is broadly described as the “interpretability” issue in AI. 

It is self-evident that interpretability is crucial for all kinds of reasons – AI has the power to cause serious harm alongside immense good. It is critical that users understand why a system is making the decisions it does. When we read and hear about ‘safety concerns’ with AI systems, interpretability and accountability are key issues.

At Ģ����Ƶ we have been working on this issue for some time – and we began way before AI systems such as generative LLM’s became fashionable. In our AI team based out of Oxford, we have been focused on the development of frameworks for “compositional models” of artificial intelligence. Our intentions and aims are to build artificial intelligence that is interpretable and accountable. We do this in part by using a type of math called “category theory” that has been used in everything from classical computer programming to neuroscience.

Category theory has proven to be a sort of “Rosetta stone”, as John Baez put it, for understanding our universe in an expansive sense – category theory is helpful for things as seemingly disparate as physics and cognition. In a very general sense, categories represent things and ways to go between things, or in other words, a general science of systems and processes. Using this basic framework to understand cognition, we can build new artificial intelligences that are more useful to us – and we can build them on quantum computers, which promise remarkable computing power.

Our AI team, led by Dr. Stephen Clark, Head of AI at Ģ����Ƶ, has published a applying these concepts to image recognition. They used their compositional quantum framework for cognition and AI to demonstrate how concepts like shape, color, size, and position can be learned by machines – including quantum computers.

“In the current environment with accountability and transparency being talked about in artificial intelligence, we have a body of research that really matters, and which will fundamentally affect the next generation of AI systems. This will happen sooner than many anticipate” said Ilyas Khan, Ģ����Ƶ’s founder.

is part of a in quantum computing and artificial intelligence, which holds great promise for our future - as the authors say, “the advantages this may bring, especially with the advent of larger, fault-tolerant quantum computers in the future, is still being worked out by the research community, but the possibilities are intriguing at worst and transformational at best.”