Cats can stare at one spot for hours.
Quantum engineers can stare at one error rate for just as long.
Now it’s your turn, stare at the grid then solve it 🐈⬛
Let us know in the comments how many words you got, and if you found the bonus word.
Watch this space tomorrow 18th November! 👀👩🍳
#QuantumComputing#CatQubits#AliceAndBob#WordSearch#PuzzleTime#DeepTech
Quantum computers aren’t that alien. 🐈⬛🛸
In the server room, a QPU is just another accelerator next to your CPUs and GPUs.
So why reinvent how we schedule and manage them?We’ve spent the past year working with the STFC Hartree Centre on QRMI, an open-source, vendor-neutral interface (alongside IBM, Pasqal and others). Implement it once, and our cat qubits work with Slurm, OpenPBS, LSF, Flux and more. And with this week’s @nvidia news on Slurm, we’re ready for what’s next. 🚀
NVIDIA’s update lets Slurm treat QPUs as first-class resources, and the QRMI community is already building the plugin for it.Our Felis cloud API already speaks QRMI, so no extra integration work needed.
Read the full story here 👉 lnkd.in/gusjarwJ
If you’re heading to @Supercomputing 26, see us run a QRMI tutorial with @IBM and @pasqal_quantum today, 13:30 CST, room W184d. 🤓
#QuantumComputing#HPC#QRMI#Slurm#CatQubits#SC26
Gartner recently released its first Emerging Market Quadrant for Quantum Computing, and Alice & Bob is recognized as a "Pioneer" and one of the most likely to bring disruption to the market 🐈⬛
A few points from the report we completely align on.
🔹No qubit technology has won yet
🔹Fault-tolerant quantum computing is expected between 2028 and 2030
🔹The market will likely consolidate around the companies that get there
Here’s where we’d add our own view 💭
The race won’t be decided by who builds a quantum computer that can “just” solve useful problems.
It will be decided by who turns that computer into an economically viable product. If a machine is too expensive to buy or to run, too difficult to build or to host, even early market traction won’t make a lasting advantage. 🏭
That’s where cat qubits make the difference, and why Alice & Bob scored so high on “potential for market disruption.” Because they exponentially correct bit-flip errors in the hardware itself, each logical qubit needs far fewer physical qubits. Fewer qubits means computers that are orders of magnitude more compact, more efficient and therefore more economical ⚡
Congratulations to everyone featured in this first edition 🥂
#QuantumComputing#CatQubits#FaultTolerance#QuantumErrorCorrection#DeepTech#Gartner
400 times the age of the universe. That’s how long today’s best supercomputers would need to factor a 2048-bit number 🔐
Then, in 1994, Peter Shor showed a quantum computer could do it efficiently, given a sufficiently large quantum computer.
Since then, the question has been how large. In our new paper, we estimate about 25,000 cat qubits could run it in about 3 days.
Dahmun Goudarzi and Vivien Londe, two of the researchers behind the paper, take you from the maths that keeps your passwords safe to the cat qubits that could change the picture. 🐈
📄 Read the paper: arxiv.org/pdf/2610.37839
📝 Read the blog post: alice-bob.com/blog/compiling…
Last week we said 22 593 cat qubits and 3.14 days. Today, the authors show their work 🐈
Our new technical blog post walks through how we compiled Shor’s algorithm for RSA-2048 with Inria Rennes, all the way from the algorithm down to every physical qubit on the chip 🧩
Inside, you’ll find 👇
🧮 The modern version of Shor’s algorithm, thirty years of optimizations in one place
🔌 Three tricks built for cat qubits (a parity-CNOT, a multi-target CNOT and active injection) that let us rebuild the adder, the operation Shor’s runs most
🗺️ A full chip layout, two layers face to face, with memory and compute zones and magic-state factories
🎯 Our assumptions, stated plainly. The headline number assumes no bit-flip on any qubit during the three-day run. Our latest preliminary results reach a bit-flip about every hour for each qubit, and the paper also works out the cost of that more conservative scenario
We also make a case for the whole field 😼 Comparing platforms only works if every estimate states its algorithm, hardware and error correction. So we propose a smaller shared milestone that uses the same building blocks as RSA-2048, one any fault-tolerant architecture could aim for first 🏁
Written by the people who did the work, and worth the read 👇
📝 Blog: alice-bob.com/blog/compiling…
📄 Paper arxiv.org/pdf/2610.37839#QuantumComputing#CatQubits#ShorsAlgorithm#FaultTolerance
Quantum computing has a size problem. We think cat qubits solve a big part of it 🐈⬛
To break RSA-2048 in about three days, we would need around 1 million standard superconducting qubits, about 33 000 m² and circa 500 MW, roughly the power consumption of Brussels 🇧🇪. With cat qubits, it's 23 000 qubits, 800 m² and 12 MW. About what an high-speed train runs on 🚄.
Why does that matter? Every qubit comes with cryogenics, wiring, control electronics and cost. A machine consuming as much as a whole city is a machine you could build, but nobody will buy. One that fits in a small building, powered by a single wind turbine, starts to look like an engineering project. 🏗️
The savings come down to physics. Cat qubits resist bit-flips by design, so error correction only has to fight phase-flips. We’ve shown this on our Boson 4 chip in 2024 and more recently on the Helium chip, where idle bit-flips times reached 1h 🧪 alice-bob.com/blog/just-out-…
This is a theory paper, such a machine still has to be built and the challenges to overcome are many. But a smaller machine is a much more realistic one, and arguably one that could be built sooner.
The world's transition to quantum-resistant cryptography has never been more urgent than now.
More on the findings this Wednesday in the blog post from the study's authors.👀 In the meantime, have a look at our infographic 👉 🖼️ alice-bob.com/applications/c…
For those of you who want to dive deeper here's the full paper: arxiv.org/pdf/2610.37839#QuantumComputing#CatQubits#ShorsAlgorithm#FaultTolerance#PostQuantumCryptography
Breaking RSA-2048 with 23,000 cat qubits. 🐈 In 3.14 days
This is the finding of our new paper, the latest resource estimate for Shor’s algorithm, the benchmark the whole field uses to compare quantum hardware 📄
The results are a record low for total quantum resources at this speed of execution, about 40 times fewer physical qubits than @GoogleQuantumAI 's Craig Gidney 2025 estimate. 🏁
RSA-2048 is the lock behind secure websites, banking and software updates 🔐 Shor's algorithm has been able to break it, on paper since 1994. Nobody has built the machine yet, but every year the estimate of how big it needs to be gets a little smaller.
In 2019, the best estimate was 20 million qubits. In 2024, ours was just under 100,000. Now 22,593, in 3.14 days. Yes, pi 🥧
Small isn't enough. Fast isn't enough. The new study optimizes for both. 😼
Bitcoin, Ethereum and most blockchains aren't secure either, about 30,000 qubits would be needed to crack a Bitcoin key in about an hour ₿ ⏳
How we got here, what it means for cryptography, and when it all converges, our CEO Théau Peronnin talks about it in the blog 👇
📝 Blog - alice-bob.com/blog/our-estim…
📄 Paper - eprint.iacr.org/2026/2665
Stay curious, more info coming soon 👀
#QuantumComputing#CatQubits#ShorsAlgorithm#PostQuantumCryptography#Cryptography#FaultTolerance#DeepTech
This Thursday something BIG yet unprecedentedly small, will come out of the lab.
To "decrypt" the riddle, you may want to play a game with us 🎲 👉 alice-bob.com/science/#shor-…
Three short puzzles, about five minutes, no physics required. You will find a rhythm, compose a score, and catch an error before it catches you… what do you think will be the big news on Thursday? Answer in the comments 👇
#quantumcomputing#FTQC#quantumerrorcorrection#crypto#shor
A small comparison of the recent QuEra + Anthropic + Harvard study and our own...
You might find it interesting. 🤖 vs 🧙
In the video: the cost of simulating Fe₂S₂ with a quantum circuit. We ran our method (doi.org/10.48550/arXiv.2610.…) on the same benchmark as their study (doi.org/10.48550/arXiv.2610.…), applying our optimizations to their circuit too.
➡️ Result: we cut TQ (the lower the cheaper) by around 1/3 vs their released circuit, by using around 2/3 fewer Toffoli gates for some extra logical qubits.
Blog post on the paper coming soon. 👀
What happens when AI enters the quantum lab, and GPUs carry the load?
At @NVIDIA GTC Berlin, our CEO and co-founder Théau Peronnin showed what that looks like in practice. AI and GPU computing now speed up every step of how we build quantum computers, from chip design to error correction:
⚡ Dynamiqs dynamiqs.org/stable/, our quantum simulation library, is integrated with NVIDIA CUDA-Q and runs on NVIDIA GPUs and simulates cat qubit physics up to ×75 faster. That means we can explore chip design ideas before we fabricate them
🔍 AI fits our measurements with ×10 less compute, checked against a digital twin of the chip
🤖 155 AI agents planned a study, our team checked the plan, and one agent ran it on our chips autonomously. A full study took a day instead of a week
🧩 In a first proof of concept with NVIDIA @NVQLink, we decoded simulated error syndromes from our LDPC-cat codes, showing how decoding can scale by becoming more efficient as several syndromes are decoded in parallel with GPUs
We also had a special visitor at the booth: @awagonfeld , NVIDIA's CMO, dropped by to see where cats and GPUs meet 💚
Thanks to everyone who came to say hi 🐾
#NVIDIAGTC#QuantumComputing#CatQubits#NVQLink
⚛️ Last week, Alice & Bob held its annual Consultative Board meeting at our Paris HQ. Raphaël Lescanne and Sébastien Jezouin met with the board to discuss our roadmap and the scientific challenges of building the first fault-tolerant quantum computer 🚀.
While there’s no instruction manual for building a quantum computer, if there were one, our board would have written it 📖:
→ Daniel Gottesman built the mathematical framework behind nearly all quantum error correction 🧮
→ David DiVincenzo defined the five criteria a machine must meet to count as a quantum computer ✅
→ John Martinis led Google’s 2019 quantum supremacy experiment and shared the 2025 Nobel Prize in Physics 🏅
→ Yasunobu Nakamura controlled the very first superconducting qubit 🥇
Their job is to help us keep strong alignment with the scientific community and progress taking place in quantum around the world 🌍.
#QuantumTechnology#QuantumErrorCorrection#Physics
Sometimes the quiet wins are the loudest. 😶🌫️
We submitted our full dossier to the @DGA sharing our first results for France's PROQCIMA quantum program. ✅
For those who may not know what PROQCIMA is, it's one of the largest public procurement programmes for quantum computers worldwide. The first phase was about demonstrating the feasibility of each candidate's technology.
▶️ Our chips are on the table, as we assembled and operated 3 QPUs on which we measured:
🔹Noise-biased CNOT gates between cat qubits
🔹A measure-and-reset sequence
🔹Multiple cycles of a three qubit repetition code
🔹Measurement of the logical error rate
🔹A non-Clifford gate on a cat qubit
Soon we'll know the verdict of our submission which is being examined as you read, but we are proud of all the progress so far... Most experiments run on Helium, our flagship QPU which tremendously improved over the summer, becoming our "error correction lab" as intended.
There's so much more coming next from our lab.
Until then, stay curious ✨
📝 Read more in the press release: alice-bob.com/newsroom/alice…