量子计算：AI拼图中的关键拼图？

#PositiveImpact Marion Laboure,Ph.D-Global Researchmarion.laboure@db.com Cassidy Ainsworth-Gracecassidy.ainsworth-grace@db.com October2023 CHARTBOOKon Quantum Computing: The Missing Piece in the AI Puzzle ? On October 24, US startup Atom Computing announced a 1,180qubit quantum processor, surpassing IBM's 433 qubit computer.This follows August restrictions by President Biden on USinvestments in Chinese tech over national security concerns aboutChina's quantum advancements. Part 1–The future of AI is quantum computing Quantum computing is the future of AI, offering faster dataprocessing and improved transparency for machine learningmodels. Part 2–The current quantum playing field While the US has historically dominated quantum technology(measured by patents) and IBM remains the company at thefrontier, other countries like China are rapidly catching up, creatinga competitive landscape. Part 3–What’s next? The widespread commercial use of quantum is not expected until2035 due to hardware and software limitations, but its importancelies in addressing the threat to encryption methods like RSA. …All in an easy-to-read PowerPoint format. WHAT IS QUANTUM COMPUTING?The technology that will solveproblems considered too complex for today’s computers Quantum computing helps solve computing problems with ahigh degree of complexityby running processessimultaneously using qubits. A qubit canstore an infinite number of values at the same time.Conceptually, it is similar to Schrödinger’s cat–bothalive and dead at the same time. ‒ Quantum computersdo not work serially(a series of things one at a time)but in parallel so that they completemultiple tasks at the same time–many multiples faster than a classical computer and today’s AI programs. ‒This means that quantum computersare more capable at solving an ‘impossible’ computer problem, such asfactorisation of a large known number into two unknown prime numbers (a virtual impossibility for today’s computers).* Classical vs quantum computing NEWS:A US startup has achieved a quantum computer with 1,180qubits On October 24, Atom Computing announced it had built a quantum computer that has 1,180 qubits. IBM had recently reached 433. ‒This follows the August 9 US executive order on Chinese quantum computing by President Biden, banning certain US investmentsinthree key industries in China: (i) advanced semiconductors; (ii) AI; and (iii) quantum computing. The restrictions are expected to comeinto force next year. ‒This is not the first US restriction on China’s quantum computing industry. In 2021, the US announced export controls on 8 Chinesequantum companies. Thisfollowed from reportsthat the industry panel ‘Xinchuangcommittee’ looked to blacklist companies thatare more than 25% foreign-owned from supply-sensitive industries. NEWS:This comes as China’s quantum computing industry becomesincreasingly sophisticated, threatening common encryption methods On October 10, a paper in Physical Review Letters announced that a ChineseJiuZhang3 quantum computer prototype solved ahighly complex mathematical problem in only a millionth of a second. The same problem would have theoretically taken thefastest supercomputer 20bn years to solve. On May 28, Xinhua reported that China has released a new generation of quantum computing cloud platform, allowingresearchers to perform complex computational tasks in the cloud. ‒The technology can pool the quantum computing power of three superconducting quantum chips featuring 136, 18 and10 qubits at a physical distance of 50km. -In December 2022, a paper published by Chinese researchers stated that they found a method to break through the mostcommon form of online encryption, the RSA algorithm.* CORPORATES:Companies are increasingly interested in quantumcomputing ‒The number of mentions of ‘quantum computing’ in corporate documents has increased from 111 in 2015 to 3,365 in 2022, anear 30-fold increase.‒IBM, Quantum Computing Inc, Alphabet and Samsung are the corporates most interested in the field, respectively, whencounting the number of corporate documents published on the topic. 1.The future of AI: QuantumComputing? You probably should have started years ago. If you wait until the last minute,when it’s clear quantum computers will be built, you will probably be too late. -Peter Shor, inventor of Shor’s algorithm, a pivotal quantum computeralgorithm NEXT GENERATION AI COMPUTING INFRASTRUCTURE:Quantumessential to address two key AI challenges: (i) compute demand; and (ii) transparency -Two key challenges facing AI technology today: -(1) Efficiently and sustainably processing increasing volumes of data-(2) Enhancing the transparency of AI processes -For some applications, switching entirely from classical to quantum computing infrastructure could address these challenges.However, in other cases, a hybrid model combining both approaches may prove more effective. -A 2021 study found that supplementing a classical commun