Thanks to recent advances in error-correcting technologies, quantum computing has never been closer to shifting into high gear.
In a statement spotted by Interesting Engineering, IBM announced that its Eagle quantum processor outperformed a traditional supercomputer when solving a complex mathematical problem. According to the firm, this is the first time that a machine of this category has proven its ability to offer precise and consistent results with more than 100 qbitsand this represents a significant step forward.
These qbits are the quantum equivalents of bits in traditional computing; in essence, they are logical subunits that store information. The main difference is that unlike classical bits which can exist as 0 or 1, qbits can exist in an intermediate state allowed by the phenomenon of quantum superposition.
A concept that doesn’t really make sense in traditional physics. But by storing complex information in the form of subtle differences in the quantum state of matter in this way, we open the way to the types of operations that traditional computers are not capable of… provided that all these qbits coexist peacefully. And that’s much easier said than done.
Making qbits collaborate, the great challenge of quantum computing
In order for them to be able to do their job, they have to be kept in this famous state of superposition—but that’s not all. The different qbits must also exhibit quantum entanglement. Very briefly, it is a state where two particles are linked by an inextricable link independent of their distance; if one of the two particles undergoes the slightest modification, the other will undergo precisely the same effects, even if it is positioned at the other end of the universe.
This is an extremely delicate mechanism, as qbits can also interact with other nearby quantum systems. Each of these interactions can disrupt the system. When they occur, the quantum entanglement so important in this case tends to collapse; we then speak of decoherence.
Quantum: researchers have doubled the lifespan of a qbit
It is therefore very difficult to maintain two particles in an entangled state. And the complexity of the problem increases exponentially with the number of qbits. This is one of the main limitations that still hinder the development of this technology. If quantum computing has not yet surpassed high-performance computing in many areas, it is largely because of this consistency problem.
Better to cure than prevent
And it is at this level that IBM researchers have just struck a blow. Instead of wasting time trying to maintain perfect entanglement between the 127 qbits of their Eagle processor, they opted for a different approach: accept that there will always be interference, and try to correct these errors rather than prevent them.
This concept is quite fashionable among quantum computing researchers. At the beginning of last year, three teams of researchers presented work where they succeeded in achieving impressive levels of precision thanks to these correction systems (see our article below). This very exploratory work was limited to tests on pairs of two qbits. But in the meantime, this approach has progressed. IBM engineers were able to apply it to the 127 qbits of the Eagle simultaneously.
By chasing errors, quantum computing is approaching a major turning point
They were thus able to carry out an extremely precise simulation of the behavior of a set of particles. The objective was to predict the physical properties of a material. This is typically the kind of problem that causes big problems for traditional computers. Even the most powerful supercomputers take a considerable amount of time to perform all these calculations. The Eagle processor, on the other hand, did very well.
Mainstream quantum computing is approaching
To analyze the results, the researchers collaborated with another team from the University of Berkeley. These researchers carried out the same operations, but on conventional supercomputers. As the size of the simulation increased, they found that the Eagle continued to deliver extremely accurate results. Supercomputers, on the other hand, eventually lost their way under the weight of this mountain of information. The researchers deduced that the correction of the errors made it possible to “ surpass the best classic simulation systems “.
” VSis the first time that a quantum system has accurately modeled a natural physical system beyond what is possible with classical approaches explains Dario Gil, vice president and director of IBM Research. ” For us, this is a very significant step in proving that quantum computers are already powerful tools, capable of solving very difficult — and perhaps even impossible — problems for traditional systems. “, he insists.
It’s all the more impressive considering that Eagle isn’t even IBM’s best quantum processor. In November 2022, the firm presented Ospreya new model to… 433 qbits (see our article). We therefore understand why Gil considers that this “ marks the beginning of a new era “.
Soon, we will finally be able to exploit the fabulous potential of this technology in a very concrete way. And with this success in hand, IBM hopes to be able to benefit its customers very quickly. Over the next year, the firm intends to offer a real cloud offer in good and due form. The press release mentions quantum computers at more than 127 qbits accessible from anywhere in the world.
A very exciting first step. Admittedly, this offer will probably take some time to mature. But it is a sign that quantum systems will soon no longer be content to imitate supercomputers. From there, it will theoretically be possible to tackle a new class of problems that were fundamentally unsolvable for traditional computers. With all that this implies for scientific research, and for our civilization in general.
