Scientists Make 'Magic State' Breakthrough After 20 Years (livescience.com) 33
An anonymous reader quotes a report from Live Science: In a world first, scientists have demonstrated an enigmatic phenomenon in quantum computing that could pave the way for fault-tolerant machines that are far more powerful than any supercomputer. The process, called "magic state distillation," was first proposed 20 years ago, but its use in logical qubits has eluded scientists ever since. It has long been considered crucial for producing the high-quality resources, known as "magic states," needed to fulfill the full potential of quantum computers. [...] Now, however, scientists with QuEra say they have demonstrated magic state distillation in practice for the first time on logical qubits. They outlined their findings in a new study published July 14 in the journal Nature.
In the study, using the Gemini neutral-atom quantum computer, the scientists distilled five imperfect magic states into a single, cleaner magic state. They performed this separately on a Distance-3 and a Distance-5 logical qubit, demonstrating that it scales with the quality of the logical qubit. "A greater distance means better logical qubits. A Distance-2, for instance, means that you can detect an error but not correct it. Distance-3 means that you can detect and correct a single error. Distance-5 would mean that you can detect and correct up to two errors, and so on, and so on," [explained Yuval Boger, chief commercial officer at QuEra who was not personally involved in the research]. "So the greater the distance, the higher fidelity of the qubit is -- and we liken it to distilling crude oil into a jet fuel."
As a result of the distillation process, the fidelity of the final magic state exceeded that of any input. This proved that fault-tolerant magic state distillation worked in practice, the scientists said. This means that a quantum computer that uses both logical qubits and high-quality magic states to run non-Clifford gates is now possible. "We're seeing sort of a shift from a few years ago," Boger said. "The challenge was: can quantum computers be built at all? Then it was: can errors be detected and corrected? Us and Google and others have shown that, yes, that can be done. Now it's about: can we make these computers truly useful? And to make one computer truly useful, other than making them larger, you want them to be able to run programs that cannot be simulated on classical computers."
In the study, using the Gemini neutral-atom quantum computer, the scientists distilled five imperfect magic states into a single, cleaner magic state. They performed this separately on a Distance-3 and a Distance-5 logical qubit, demonstrating that it scales with the quality of the logical qubit. "A greater distance means better logical qubits. A Distance-2, for instance, means that you can detect an error but not correct it. Distance-3 means that you can detect and correct a single error. Distance-5 would mean that you can detect and correct up to two errors, and so on, and so on," [explained Yuval Boger, chief commercial officer at QuEra who was not personally involved in the research]. "So the greater the distance, the higher fidelity of the qubit is -- and we liken it to distilling crude oil into a jet fuel."
As a result of the distillation process, the fidelity of the final magic state exceeded that of any input. This proved that fault-tolerant magic state distillation worked in practice, the scientists said. This means that a quantum computer that uses both logical qubits and high-quality magic states to run non-Clifford gates is now possible. "We're seeing sort of a shift from a few years ago," Boger said. "The challenge was: can quantum computers be built at all? Then it was: can errors be detected and corrected? Us and Google and others have shown that, yes, that can be done. Now it's about: can we make these computers truly useful? And to make one computer truly useful, other than making them larger, you want them to be able to run programs that cannot be simulated on classical computers."
Yep, suuure... (Score:3)
Unless this gets independently reproduced, I am not believing one word of that claim.
Re: (Score:2)
Very cool though if confirmed to be true.
Re: (Score:3)
If confirmed yes, before that hot air. And it may not mean what the story claims it means either, even if confirmed.
Re:Yep, suuure... (Score:4, Interesting)
further, if it is confirmed then there's still a potentially decades long process for them to actually be useful... regardless of the rhetoric in the article.
and they still wont be able to accurately predict the stock market, because some types of quantum states are stochastic and functionally reliant on external inputs not just some rationalisation about states etc etc.
People these days with hyping new physics... they seem to think that every time we make any technological progress at all they should be able to try to ignore conservation, entropy and SNR, but at the same time they seem to think only the stuff with a press release happens at all.
Re: (Score:2)
Oh so... what's going on here, if i'm reading it correctly... is they're limited to clifford gates and the stabilization they're talking about might just be combining the real output from an actual qbit computation with a classical version simulated on traditional x86 hardware.
This is notably less impressive than it could be even if I've got that last bit wrong, because according to the gottesman-knill theorem... anything that can be built using only those gates in a quantum system can be simulated at polyn
Re: (Score:2)
Interesting. But p-time is probably somewhat misleading. In practice, it matters very much what the exponents are. Still, I am there with you, being prematurely impressed by this result is not rational.
Re: Yep, suuure... (Score:2)
How much of the universe, from entanglement to the predicted zero point energy vs. actual, is rational? Remember how Hippasus is rumored to have been killed for discovering irrational numbers? Would you have supported that?
Re: (Score:2)
People these days with hyping new physics... they seem to think that every time we make any technological progress at all they should be able to try to ignore conservation, entropy and SNR, but at the same time they seem to think only the stuff with a press release happens at all.
Indeed. People look for magic, but education is bad enough that they do not understand there is very little spaces left where magic could potentially be found.
Re: (Score:2)
How did the stock market get into this discussion?
Re: Yep, suuure... (Score:5, Funny)
Re: (Score:2)
I will accept magic if solidly proven and reproducibly demonstrated. Not before. You know, doing actual Science. (No, Science does _not_ rule out magic. It just places high requirements on the proof of its existence.)
Re: Yep, suuure... (Score:2)
How is your position different from an epicyclist who, instead of improving the sensitivity of his sensors, simply declared Aristarchus' heliocentric theory disproved because star parallax could not be measured?
Re: (Score:2)
Is this a serious question? I am merely saying "I want proof" above the very basic level.
Re: Yep, suuure... (Score:2)
Why can't I be perfectly serious as I ask how you know you aren't falling into the same trap as the epicyclists who also demanded proof that they were wrong, and the proofs supported them? If I supported heliocentricism because it made sense without proof, would you be justified in calling me a crank?
Re: (Score:2)
In science, it is always valid to ask for proof. In fact, if you do not, you are not doing science. If the claims are more out-there or there is a reasonable suspicion of a conflict of interest (as there is here), it is always acceptable to ask for better proof.
And if you "support" something, then you are square in the area of religion or politics. Facts do not need "support". Made up stuff does.
Re: Yep, suuure... (Score:2)
Do you see the problem with proof, though? How come geologists look at Iceland or the coasts of Africa and South America today and see unmistakeable proof of continental drift, yet a mere few decades ago the same proof was mocked mercilessly by stalwarts such as yourself?
A little confused on "Distance". (Score:2)
Is "Distance" as used here:
a. A measure of how physically far apart the Qubits were (N nM, N*2 nM, etc.): a specified Distance that drives the results they claim
or
b. A characterization of how the experimenters did at a given distance: "We successfully reconstructed the magic state at Distance 5 and Distance 2"
I can see on some level, these are similar if not the same (same kind of problem as "what does 20% probability of precipitation mean?": depends how you look at the scenario).
But in a lexical sense, wha
Re: A little confused on "Distance". (Score:2)
Each logical qubits is composed of multiple physical qubits required for error tolerance. I believe distance in this context is how many of those physical qubits needs to fail before the logical qubit is corrupted.
Re:A little confused on "Distance". (Score:4, Informative)
Hamming distance is the term you're looking for.
Re:A little confused on "Distance". (Score:4, Informative)
Great question—distance here doesn't mean spatial separation. It refers to quantum error correction, specifically the code distance in a quantum error-correcting code (like a surface code or stabilizer code). It’s directly tied to how many physical qubit errors the system can detect and correct. If you've had some undergraduate CS courses, think Hamming distance, but for quantum computation. In classical coding theory, Hamming distance measures how many bit flips are needed to turn one valid codeword into another. A higher distance means better error tolerance.
In quantum codes, a code with distance d means the system can detect up to d–1 errors and correct up to floor((d–1)/2) errors across physical qubits. So a “distance-5 logical qubit” is a logical qubit encoded across many physical qubits, with enough redundancy to correct 2 errors. That’s a big deal when you're building magic states—special resource states needed to implement non-Clifford gates like the T-gate, which are essential for universal quantum computing.
Clifford gates (like Hadamard and CNOT) are easy to do fault-tolerantly. But the T-gate isn’t. To work around this, we inject carefully prepared magic states into the circuit. Problem is, those states are fragile—and encoding them in distance-5 logical qubits gives them a fighting chance at surviving real-world noise.
So, in a bucket, "Distance" = error-correcting strength, not spatial distance. It’s measured like Hamming distance in classical computing, but adapted for quantum codes. Bigger distance = better fault tolerance = more reliable magic states. Magic states are how we implement T-gates—and without T-gates, a quantum computer is stuck in the Clifford-gate only zone, which means they are no faster than a classical computer at solving problems -- no quantum speedup, no Shor’s algorithm, no magic. T-gates (via magic states) are what push the system into the fully universal quantum computing regime—where it can solve problems that classical systems fundamentally can’t keep up with.
Hope that helps clear it up.
Re: (Score:2)
A most excellent reply, thanks!
Um ... (Score:2)
Re: (Score:2)
First it was spooky, then dark, now it's magic... Terms that don't instill confidence that even the experts know what is going on or how to get anywhere.
Re: Um ... (Score:2)
What if it's just mood affiliation all the way down to quark level ("charm" ...)?
Convergence, only arcane (Score:4, Interesting)
Re: (Score:2)
we are still in the 40’s to having a working digital personal computer.
I think you are being optimistic here.
Nonsense? (Score:2)
Re: (Score:2)
A befuddled investor is still an investor, but one that sees clearly that you aren't getting anywhere in their lifetime may no longer be an investor.
Re: Nonsense? (Score:2)
Why are those investors simultaneously smarter than you for having money and dumber than you for throwing it away on projects that befuddle you?
"Us and Google" (Score:4, Insightful)
Re: (Score:2)