Chinese researchers have presented a very powerful new algorithm on paper, but the American contingent is skeptical.
Alongside artificial intelligence, quantum computing is regularly cited among the technologies that could completely transform the face of humanity.
The power of this technology comes mainly from the fact that it can store complex information in the form of subtle differences in the state of matter. Indeed, quantum computers work thanks to logical sub-units different from the bits found in traditional computers; here, we are talking about qubits.
Very vulgarly, their particularity is that they can exist in several distinct states simultaneously, unlike conventional bits which take either the value 1 or 0. Without going into detail, what is important to remember is that is that this architecture makes it possible to deal with certain problems immensely faster than with a lambda PC.
Current encryption algorithms are on borrowed time
When they come of age, quantum computers will pave the way for dramatic advances. And this concerns almost all possible and imaginable disciplines. But there is also a flip side. With a machine of this kind, it would theoretically be possible to pulverize all modern encryption algorithms. Including the strongest.
A pretty terrifying prospect. Because today, these algorithms make it possible to protect lots of excessively important and sensitive infrastructures and services. A simple and very evocative example: with a mature quantum computer available, the concept of password as we know it today is in great danger of being shattered.
Current algorithms should therefore be replaced very quickly by other encryption systems capable of withstanding the onslaught of quantum computers. There are already a few systems that could take over. But making such a drastic transition from a global infrastructure in a rush would surely be an exceedingly perilous leap of the angel.
The good news is that we are still quite a long way from that deadline. Specialists traditionally consider that it would take a machine with several million qubits to dismantle the most advanced encryption algorithms. And for now, the most powerful quantum computers still have to make do with a few hundred qubits. Even Osprey, the record quantum processor unveiled in the fall of 2022 by IBM, only offers 433 (see our article).
Or at least that’s what everyone thought. Because a big surprise awaited the scientific community last weekend. Specialists were caught off guard by the latest work from Tsinghua University in China.
A fairly revolutionary algorithm… on paper
According to the South China Morning Post, a team led by Professor Long Guilu presented a new algorithm based on Shor’s algorithm. The latter is a mathematical tool proposed by the physicist of the same name in 1994. In theory, it would make a quantum computer exceptionally fast when it comes to breaking a cipher.
The researchers explain that their new algorithm, called SQIF (Sublinear Resource Quantum Integer Factorization), would drastically reduce the number of qubits needed to arrive at practical applications. And we’re not talking about a small reduction; they claim that their algorithm would allow a 372-qubit machine to crack RSA-2048!
For reference, this is one of the strongest security standards in the digital world. It should several hundred billion years to a standard PC to find the 2048 elements of the key. If so, it would be a sacred revolution in the world of computer securityat all levels.
However, in science, when a team presents such spectacular assertions, it is better that the paper rests on exceptionally solid theoretical bases. And this is where the problem starts to get both very meandering, but also very interesting. Indeed, the demonstration part did not really live up to the initial claims.
As a proof of concept, the researchers cracked a 48-bit encryption key with a tiny quantum computer of just 10 qubits. A result certainly impressive, but still very far from threatening RSA-2048. Especially since researchers have no proposed any track that would allow to make the link between their results and this objective. And that point in particular raised a few eyebrows.
The first disturbing element is that it is not no secret-defense classified work. However, traditionally, the great powers like China tend to lay a leaden screed on all the technologies which could bring them a decisive strategic advantage. This apparent inconsistency has alerted the scientific community. Some researchers have serious doubts about the reliability of the results.
We can cite in particular Scott Aaronson, a great American specialist in algorithms and encryption. He maintains a fascinating blog where he often delivers very well vulgarized analyzes on problems of this kind. A few years ago, he notably gave birth to a text where he explained how solving one of the Millennium Problems would make it possible to get hold of all the existing Bitcoins.
Solving this mathematical problem would make it possible to monopolize all the existing bitcoins
And the least we can say is that he is particularly skeptical of the claims of the Chinese team. “ It seems to me that it would take a miracle for this approach to bring any benefit “, he says in a blog post. ” Rarely have I seen such an actively misleading paper in quantum computing research in 25 years… and I have seen many. »
And he’s not the only one to think so. Peter Shor, the author of the original algorithm that served as the basis for SQIF, also expressed some reservations spotted by SCMP. ” There are apparently several possible problems with this paper “, he replied to one of his colleagues who accused the Chinese team of having ” omitted crucial technical details “. The formulation may seem very sober, but in the academic world, it is a rather frontal attack.
Looking through these diverse opinions, however, one notices a striking trend. It turns out that most of the outspoken criticism comes from the same place, namely the United States. However, as soon as a question involves both China and Uncle Sam, everything immediately becomes more complicated.
Indeed, the two superpowers are currently entangled in a kind of pseudo-cold war. They actively seek to put a spoke in the wheels. This is particularly evident on the American side; the Biden administration, for example, has launched a huge subsidy plan worth several hundred billion whose assumed and claimed objective is to ” counter china (see our article).
We find ourselves in a somewhat comical situation. Because when the United States talks about China and vice versa, it is often difficult to determine what is legitimate propaganda or criticism. Especially since the two giants make it a point of honor to cover their tracks.
Nauseating artistic blur
Would these announcements be a vast political smokescreen intended to panic the American administration ? This seems unlikely, as Long is a rather reputable researcher. Another recognized specialist in Chinese encryption, interviewed on condition of anonymity by the SCMP, in any case wanted to defend it. ” Professor Long conducts solid research and has always been discreet “, explains the interested party. ” He’s not going to risk his academic reputation for political reasons. »
So is it a spectacular scientific advance that Americans would refuse to recognize out of pride ? It also seems far-fetched, because even though they are mostly American, the reviews still come from referenced specialists.
So is it just a somewhat wobbly scientific publication as there are already millions of them, but which would have caused strong reactions only because of the subject of the study ? This seems to be the most satisfactory explanation for this scientific-political imbroglio. But for the moment, it is still impossible to be sure.
Indeed, the paper has not yet passed the stage of peer review. This is an essential step in the scientific publication process where other specialists must verify the validity of a study before it is published. It will therefore be necessary to wait for the paper to finish its academic course.
In any case, this case clearly shows that this theme is beginning to generate a certain excitement among the major world powers; it gives us a a taste of the friction and major upheavals that could emerge when a country finally masters this revolutionary technology. It will therefore be interesting to observe how the Chinese and American governments will manage this transition, and what place Europe will have in this story.
The text of the pre-publication study is available here.