Lead Systems Engineer

Quantum, AI Photonics: A New Computing RevolutionComputing and information technologies are undergoing several technological revolutions at once: the rise of AI, the emergence of quantum computing, and the turn to photonics to overcome the limitations of classic silicon computing.So far, each of these new sectors has mostly worked in isolated silos: AI training and computing is performed on classical silicon chips, quantum computing is looking to improve its technology until it can find a practical use, and photonics technology is still experimenting on designs and applications.Maybe unsurprisingly, it could be that it is by merging these fields together that new possibilities emerge. It seems that quantum computing might just have found a practical use case, and would not even need further improvement before being useful.Researchers at the Vienna Center for Quantum Science and Technology (VCQ) (Austria), the Politecnico di Milano (Italy), Consiglio Nazionale delle Ricerche (IFN-CNR) (Italy), and the company Quantinuum (UK) have found that existing quantum computers could outperform classical computer in AI training, using a photonic processor.They published their results in Nature Photonics1, under the title “Experimental quantum-enhanced kernel-based machine learning on a photonic processor”.Why AI Training and Quantum Computing Face LimitsAI Training’s Soaring Costs and Energy DemandsRecently, AI technology has made tremendous progress. This was however achieved only through the use of mind-boggling amounts of computing power, consuming tens of billions of dollars in chips and electricity.Certainly, some progress can be made in efficiency, as demonstrated by DeepSeek AI trained with ultra-low cost in both computing and money, beating its Western competitors by an order of magnitude. But still, at the end of the day, software improvement will only go so far in making AI training less compute and power-intensive.Quantum Computing’s Scalability and Noise ChallengesMeanwhile, Quantum computing is a promising technology, but until now suffering from a fatal flaw. The extremely fragile state of matter needing to be maintained for quantum computing to work means it is both expensive and not very scalable.It also means that the results obtained are “noisy”, with regular errors, delays, and unreliable results.Here too, innovations could mean that either a network of smaller quantum computers or a new hardware design architecture, using a new state of matter called topoconductors, allowing for scalability, could solve the problem.Until it is confirmed, this has nevertheless put into question the relevance of quantum computing, still a technology in search of a practical use case that makes sense economically.Quantum-Enhanced Kernel Methods for AIHow Quantum Kernels Add Dimensional Power to Machine LearningKernel methods are widely used tools in machine learning, and use a mathematical method of adding dimensions to a dataset for better identification of hidden patterns.This, of course, involves quite complex mathematics, which will be mostly understandable only for a limited set of specialists already working in this field. You can see a visual representation of how it works in this video:And such complex calculations could be a perfect fit for the unique capacities of quantum computers.Photonic Processors Meet Quantum Kernels for AIAn integrated photonic processor, created via femtosecond laser writing on a borosilicate glass substrate, was used for this experiment to encode the data into a state that can be treated by a quantum computer.This way, kernels displaying quantum interference were used for the calculation and compared to classical methods.Experimental Results: Quantum vs

Decrypt’s Art, Fashion, and Entertainment Hub. Discover SCENEGoogle just dropped a new research paper, and Bitcoin maxis may want to do some quick math. The tech giant's quantum team found that breaking the RSA encryption protecting everything from your bank account to your Bitcoin wallet might need 20 times fewer quantum resources than previously estimated.“Planning the transition to quantum-safe cryptosystems requires understanding the cost of quantum attacks on vulnerable cryptosystems,” Google Quantum Researcher Craig Gidney wrote. “In Gidney+Ekerå 2019, I co-published an estimate stating that 2048 bit RSA integers could be factored in eight hours by a quantum computer with 20 million noisy qubits. In this paper, I substantially reduce the number of qubits required.”“I estimate that a 2048 bit RSA integer could be factored in less than a week by a quantum computer with less than a million noisy qubits,” Gidney argued."This is a 20-fold decrease in the number of qubits from our previous estimate," the Google researcher said in an official blog post.Image: GoogleBut it’s not like it’s going to happen anytime soon. For context, IBM's Condor (the most powerful quantum computer to date) tops out at 1,121 qubits while Google's own Sycamore runs on 53