Good Morning,
I have been thinking about Microsoft’s push of their topological qubit approach for the past few days. When Google or Microsoft claim breakthroughs in something they aren’t known for, we need to be skeptical.
Microsoft’s latest work, published in Nature Magazine, introduces Majorana-1, a 𝗻𝗲𝘄 𝗾𝘂𝗮𝗻𝘁𝘂𝗺 𝗽𝗿𝗼𝗰𝗲𝘀𝘀𝗼𝗿 𝗱𝗲𝘀𝗶𝗴𝗻 aimed at leveraging 𝘁𝗼𝗽𝗼𝗹𝗼𝗴𝗶𝗰𝗮𝗹 𝗾𝘂𝗯𝗶𝘁𝘀. Their study discusses a single-shot interferometric measurement of fermion parity—a fundamental step toward measurement-based quantum computing.
Satya going on Dwarkesh’s podcast is going a bit too far! Microsoft is not an AGI company nor is its approach anywhere close to a million qubits.
The PR is on the borderline of BigTech misinformation, and given that Microsoft owns LinkedIn, this is a major problem for real journalism, facts or accountability journalism and media with a balanced perspective. If the Billionaire class controls social media like we have seen recently with X, LinkedIn and Threads, we aren’t getting facts and real journalism. Most enthusiasts or evangelists simply spread the misinformation, portraying it as factual.
There are major gaps in how Microsoft is framing the enormity of the challenge, mostly to make themselves look like a leader. The reality is it’s much more complicated, and the world still doesn’t know which qubit approach is the best for scaling Quantum computers to 1 million qubits. There’s a fairly high chance it’s not the topological qubit, the approach Microsoft has taken.
Gamechanger from Microsoft?
Good on Dwarkesh for doing lucrative PR for Microsoft. However, that’s his job as a YouTuber. I get it. He wants to profit and Microsoft wants to reach a bigger audience.
Every week I’m reading science news about a “Quantum breakthrough”, it’s a title that garners clicks and yet illustrates how absurdly far we are from realizing real quantum computers. Google or Microsoft’s Quantum chips don’t really move civilization forwards. There’s not just a lot of unknowns here, there’s a lot of exaggeration throughout the years by IBM, Google and Microsoft in their approaches to Quantum.
Microsoft Hasn’t Made Tangible Headway
While the Microsoft announcement shows work that does not demonstrate a functional Majorana qubit, it does introduce an interesting new architecture, according to Cecile M. Perrault of Alice & Bob.
Quantum computing has no lack of possible and speculative qubit architectures to scale Quantum computers into reality.
Major Qubit Approaches
As of 2025 we have no idea which qubit approach will reach the goal, if any of the current ones under experimentation and with R&D. We do however have an idea of which might be most likely to achieve QEC and scalable quantum computers.
Microsoft’s topological qubits aren’t among the common sense leaders by the way.
Cecile continues:
𝗛𝗼𝘄 𝗗𝗼𝗲𝘀 𝗧𝗵𝗶𝘀 𝗖𝗼𝗺𝗽𝗮𝗿𝗲 𝘁𝗼 𝗢𝘁𝗵𝗲𝗿 𝗤𝘂𝗯𝗶𝘁𝘀?
Quantum computing is evolving with multiple approaches, each with its own strengths:
• 𝗦𝘂𝗽𝗲𝗿𝗰𝗼𝗻𝗱𝘂𝗰𝘁𝗶𝗻𝗴 𝗾𝘂𝗯𝗶𝘁𝘀 (IBM , Google , Amazon Web Services (AWS) , Alice & Bob, Rigetti Computing ) – The most established technology, with the most progress on error correction.
• 𝗧𝗿𝗮𝗽𝗽𝗲𝗱 𝗶𝗼𝗻𝘀 (IonQ , Quantinuum ) – Highly coherent but the path to scaling is challenging.
• 𝗦𝗽𝗶𝗻 𝗾𝘂𝗯𝗶𝘁𝘀 (Intel Corporation , C12 , Quobly ) – Semiconductor-based, promising integration with classical computing.
• 𝗣𝗵𝗼𝘁𝗼𝗻𝗶𝗰 𝗾𝘂𝗯𝗶𝘁𝘀 (PsiQuantum , Xanadu , Quandela ) – Using light for quantum processing and long-distance networking.
• 𝗡𝗲𝘂𝘁𝗿𝗮𝗹 𝗮𝘁𝗼𝗺𝘀 (Pasqal , QuEra Computing Inc. ) – Using arrays of cold atoms as scalable qubits with long coherence times.
In the superconducting category, some are tailored for efficient error correction:
• 𝗖𝗮𝘁 𝗾𝘂𝗯𝗶𝘁𝘀 (Alice & Bob, Amazon Web Services (AWS) ) – a sub-class of superconducting qubits with passive error protection
• 𝗧𝗼𝗽𝗼𝗹𝗼𝗴𝗶𝗰𝗮𝗹 𝗾𝘂𝗯𝗶𝘁𝘀 (Microsoft ) – Another subclass of superconducting qubits, theoretically robust but experimentally challenging
• 𝗙𝗹𝘂𝘅𝗼𝗻𝗶𝘂𝗺 𝗾𝘂𝗯𝗶𝘁𝘀 ( Atlantic Quantum )– A type of superconducting qubit with improved coherence properties
What I think
I certainly believe that Neutral atoms or photonic qubits are much better prospects for the “right qubit approach” to get us to scalable Quantum computers than Microsoft’s moonshot here.
Microsoft promoting their architecture as a “breakthrough” only reinforces my feelings on this.
How to Summarize Qubit Approaches:
Let’s try again to simply summarize what these different approaches do.
🟢 Topological Qubits
Topological qubits are a novel approach aimed at providing inherent error resistance through their unique braiding properties, suggesting a robust path forward for fault-tolerant quantum computing that reduces the need for extensive quantum error correction mechanisms.
🔵 Superconducting Qubits
Superconducting qubits work by using superconducting materials to create circuits that can oscillate between quantum states. They are currently one of the most widely used types in quantum computers, with major companies like IBM and Google utilizing them.
🟣 Trapped Ion Qubits
In trapped ion quantum computing, individual ions are manipulated using lasers. The ions are held in place using electromagnetic fields, and their internal quantum states serve as qubits. This method allows for high precision and coherence but can be technologically challenging to scale.
🟠 Photonic Qubits
Photonic qubits use properties of photons, such as their polarization or phase, to represent quantum information. This approach benefits from robustness against environmental interference and is particularly promising for quantum communication and networking applications.
🔴 Neutral Atom Qubits
Neutral atom qubits involve trapping and manipulating neutral atoms using laser light. They allow for scalable architectures due to the ability to control interactions between atoms at a distance.
🟡 Quantum Dots
Quantum dots are semiconductor-based qubits that leverage the quantum mechanical properties of confined electron states. They can be integrated with existing semiconductor technology, presenting a potential pathway for future quantum computing devices.
🔘 Defect-Based Qubits
Defect-based qubits use imperfections in solids, such as vacancies in diamond lattices (e.g., nitrogen-vacancy centers), to represent qubits. These defects can be manipulated optically and are promising due to their stability and coherence times.
Startups, Innovation and Reality
Most of these approaches will be dead-ends, a lot of the current leading Quantum startups in their prototypes won’t exist in the 2030s. That’s just the reality of Quantum startups pretending like they are “one breakthrough away” from Quantum computers with millions of qubits, but the progress in reality is actually painfully slow.
Microsoft promoting their moonshot as dogma isn’t doing anyone any favors, unless you happen to own a lot of fairly poor Quantum computing penny stocks. In this Newsletter I have talked repeatedly about whom I am referring to.
Microsoft’s Topological architecture is interesting, don’t’ get me wrong, but it’s not a breakthrough. Just like Google’s PR in December, 2024 wasn’t real or tangible. These are absurd exaggerations to anyone in the field, who obviously need to keep their mouths shut. They have financial incentives to exaggerate the capabilities of the work they are doing.
Read the transcript of Satya Nadella’s YouTube appearances here. But don’t take what he claims at face value. It’s simply PR.
Cecile concludes:
“Scaling up quantum computing requires 𝗾𝘂𝗯𝗶𝘁𝘀 𝘁𝗵𝗮𝘁 𝗰𝗮𝗻 𝗵𝗮𝗻𝗱𝗹𝗲 𝗲𝗿𝗿𝗼𝗿𝘀 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁𝗹𝘆. Whether through Majorana qubits, cat qubits, or neutral atoms, reducing the need for excessive error correction is key to making quantum computing practical.”
Microsoft’s February 19th PR announcement containing few technical details.
Satya comparing their speculative Topological qubit architecture to the history of the transistor is ludicrous. It’s hard to listen to, because frankly, it is a form of misinformation.
Unfortunately in 2025, whether its Google DeepMind, Microsoft or OpenAI, the PR is a kind of illegal lobbying where they have too much control over social media pushing somewhat fairly false spins on innovation and research. It doesn’t hold up to scrutiny, facts or even rudimentary investigations.
As of 2024, the ratio of public relations (PR) professionals to journalists in the United States has reached approximately six to one.
Unfortunately as democracy and capitalism unravel, Silicon VAlley have irrevocably hurt real journalism and what’s left of the the media establishment is itself Billionaire backed.
It was the job of Catherine Bolgar to portray Microsoft as a winner. Their approach is a decent Moonshot, where “the breakthrough” required developing an entirely new materials stack made of indium arsenide and aluminum, much of which Microsoft designed and fabricated atom by atom.
Catherine on her website says, she has written over a thousand articles on nearly every topic imaginable. Yet indeed, it is a work of the imagination too. If you don’t have technical details to present and are employing a PR writer of this variety, what reasonable conclusion should we draw?
After all, topological qubits are themselves speculative.
Majorana 1 chip Promo Video
This is the lowest grade of vague quantum computing rhetoric that we’ve been hearing for the last decade or more. Spin out your Moonshot, IPO it, do something! Don’t just put out weak PR like this. Grow some balls Microsoft.
What is the Majorana 1 chip?
So why is making a chip around it deceptive?
It is touted as the world’s first quantum processor powered by a Topological Core architecture that allows it to potentially scale.
The Majorana 1 chip is based on a new architectural concept leveraging a special class of materials known as topological superconductors or topoconductors. These materials are designed to observe and control Majorana particles, which are exotic quasiparticles theorized to exist under specific conditions.
People like Matthias Troyer have been at Microsoft for over 19 years. Can you guess after two decades where the real state of this technology is in terms of commercializing into a real product? I dare you to take a wild guess.
Microsoft likely didn’t get DARPA partnerships on merit, you see. Microsoft is a Cloud provider to the DoD. Increasingly moving into Poland you understand what this means for European foreign affairs and politics. Microsoft has established itself as a key cloud service provider for the United States Department of Defense (DoD) through its Azure platform.
Microsoft’s topological qubit architecture is so out there, it’s highly likely not even among the top two dozens companies in Quantum computing. Pretending you are, is just not going to cut it in such a competitive world.
Microsoft faced a significant setback when it had to retract a controversial research paper published in 2018, related to quantum computing, which was featured in the journal Nature. They have a history like Google does, of exaggerating their capabilities.
What are Majorana fermions?
Microsoft's Majorana fermions refer to a special class of particles that are pivotal in the company's development of quantum computing technology. The concept of Majorana fermions originates from theoretical physics, where they are defined as particles that are their own antiparticles. These particles were first proposed by Italian physicist Ettore Majorana in 1937, and their unique properties inspire new methods for storing and processing quantum information.
It’s a very cool idea! If Topological qubit architectures scale, it’s not likely to be the work under Microsoft that does it or achieves it. 1937! So an idea that’s almost 100 years old.
Majorana fermions are utilized to create topological qubits, which promise greater stability and lower error rates compared to traditional qubit designs.
Research and experimentation around creating and measuring Majorana fermions have been ongoing and it’s not actually clear where Microsoft is along the journey.
What is DARPA's Quantum Benchmark Initiative?
The Quantum Benchmark Initiative (QBI) is an innovative program initiated by DARPA (Defense Advanced Research Projects Agency) aimed at assessing the viability and long-term utility of quantum computing technologies. Its primary focus is to create new benchmarks that will quantitatively measure the progress of quantum computers. This initiative seeks to establish reliable metrics for evaluating the performance and potential applications of quantum computing systems in various fields.
What is more likely is this program yielded some good results. It’s lucrative for Microsoft to work with DARPA. However they aren’t transparent about that either.
The exact financial details regarding Microsoft's collaboration with DARPA on quantum computing, such as the total funding amount or specific financial arrangements, have not been publicly disclosed.
DARPA’s QBI program sounds fairly speculative as well. Their goal is to do it by 2033, eight full years away. Just to verify which approach might be the best! That’s not very reassuring.
The Quantum Benchmarking Initiative is an expansion of the existing DARPA Underexplored Systems for Utility-Scale Quantum Computing (US2QC).
Microsoft’s Moonshot Architecture
It’s highly unlikely Microsoft’s topological architecture will build Quantum computers with over a million qubits in the next 10 years, perhaps not in the next 20 years. But another qubit approach is likely to make significant progress (i.e. logical qubits) in this period. Also have to add, fairly probable Microsoft’s topological approach never comes to practical fruition of scale.
Microsoft has spent 17 years researching a new material and architecture for quantum computing. But there’s no actual evidence it has made good progress.
Microsoft Quantum is a branch of Microsoft focused on developing quantum computing technologies and solutions. There’s not much grounding in reality that Microsoft Azure Quantum is a leader in the field. If anything, I have more confidence in Amazon here via Amazon Braket. What Nvidia is doing in Quantum is also more realistic.
I have very little faith in Microsoft’s two decades in Quantum research. Nor the publications that do PR for them that are supposed to be real journalists. Do I believe that Majorana 1 could fit a million qubits onto a chip comparable in size to modern CPUs? Not in the slightest. Do I believe this is a Quantum milestone like dozens of Evangelists on LinkedIn parroting the PR? Not at all.
Is it entertaining reading geared to promote National defense funding into Quantum and boost speculative Quantum penny stocks? Perhaps if you are gullible to American BigTech lobbying.
They are using Ettore Majorana’s name, probably in vain. It’s very much a story for the NYT, to get behind, who also do PR for BigTech. I’m saying this while we are in a Quantum winter, with funding drying up but “breakthroughs” arriving on a weekly basis in pop-science magazines.
What is the Quantum Winter?
As an emerging tech analyst I’m more bullish about the progress in robotics and humanoid general purpose robots, but not Quantum. "Quantum winter" refers to a period of stagnation in the development and investment in quantum computing technology, characterized by decreased interest and financial backing.
Microsoft would be among my last choices to bet on for the future of Quantum. Satya Nadella has a history of some pretty exaggerated claims when Generative AI first came out. When you have to retract a paper from Nature, it tells you all you need to know.
Qubits can be created in different ways, each with advantages and disadvantages. It might take another 20 years to even build a real quantum computer with trillions of calculations and over a million qubits. Topological architectures are a moonshot, at best.
CEO of Nvidia on Quantum’s Future
Recently, Jensen Huang, the CEO of Nvidia, made significant remarks regarding the future of quantum computing, suggesting that useful quantum computers may still be 15 to 30 years away. During various presentations and analyst calls, he emphasized the challenges and limitations currently facing the technology, asserting that the practical application of quantum computing is likely two decades away.
I concur with that realty. Who are you going to trust on this, Satya Nadella or Jensen Huang?
Huang's comments came during Nvidia's analyst day at CES 2025.
For the record I like Krysta Svore, I think she also made some nice videos.
But do I think Microsoft has made “the world’s first Quantum Processing Unit (QPU) powered by topological qubits” that has a viable future? No I don’t at all.
This is the problem with R&D in-house at a firm like Microsoft. They are paid to exaggerate the success of their work forcing engineers and scientists into petty PR spinners.
“While the built-in protection of topological qubits simplifies quantum error correction (QEC), we still need reliable QEC to get us to a fault-tolerant quantum supercomputer. Today, our team also published a roadmap to show our path from single-qubit devices to arrays that enable quantum error correction where our custom QEC codes can reduce overhead roughly tenfold compared to the previous state of the art approach. This dramatic reduction means that our scalable system can be built from fewer physical qubits and has the potential to run at a faster clock speed.”
They don’t have QEC, all they have is a roadmap. Big surprise.
Plenty of Quantum startups have roadmaps too, and papers and funding.
Are engineered Qubits the future? We actually don’t know.
Is Microsoft a potential leader in topological architectures? Perhaps.
What did Microsoft actually provide? Well, a roadmap, kind of.
It’s impossible to verify the integrity of the claims they have made with the lack of data and technical details they have provided. In the PR world startups typically do this when they don’t have anything tangible. Having a picture of a new Quantum chip prototype, means nothing in such a context.
The use of Majorana fermions is fascinating, but for practical Quantum computers, it’s probably a dead end.
If you were a top physicist or Quantum researcher, it’s fairly unlikely you’d want to work at Microsoft. Building or saying that you have built the world’s first fault-tolerant prototype (FTP) based on topological qubits is a nice ploy, especially considering how badly Microsoft is failing in Generative AI. Is fairly distracting from your R&D failures, which were fairly expected given you funded your main competitor (OpenAI).
For Microsoft as an entity of research, there are bigger issues at play here.
Google’s Willow Chip (first image) or Majorana 1 Chip (second image) gives the impression you are a leader in complicated future technologies.
It doesn’t take much digging into the facts to see how far this is from the reality of innovation and progress in the space. The Google PR (late 2024) had a huge impact on the stock market, and this chip story is even more speculative.
Current estimates suggest that there are roughly 6.2 PR professionals for each journalist, which underscores the challenges faced by journalists in an environment increasingly saturated with public relations activities and a social media platforms run by BigTech and Billionaire interests. The ratio of PR to investigative journalists by the end of the 2020s will be closer to 10:1.
In December 2022, the DoD announced its decision to utilize Microsoft Azure as part of its Joint Warfighting Cloud Capability (JWCC), a multi-cloud initiative designed to enhance operational capabilities across the military services. This contract, valued at approximately $9 billion.
Don’t be surprised if the Topological architecture and news about the Majorana 1 chip fades into the white noise of the Quantum winter. Microsoft and Google are spending so much on Generative AI and AI Infrastructure they certainly don’t have decent R&D budgets for quantum computing. Of course this is what makes all of this PR so laughable.
That’s after all, why SandboxAQ was spun out of Google.
Share this post