On the afternoon of the 6th, hundreds of audience members filled the Korea Advanced Institute of Science and Technology (KAIST) Academic Cultural Center in Daejeon. An academic conference was held in collaboration with the prestigious journal Nature, organized by the Department of Materials Science and Engineering at KAIST. The theme of the conference was 'New materials for artificial intelligence and artificial intelligence for new materials,' featuring scholars from various fields utilizing AI both domestically and internationally.
In particular, two corporations caught the attendees' attention. Mark Lam, the head of design automation at NVIDIA, a company that has shaken the global semiconductor market with graphic processing units (GPUs), was one of the speakers. It is unusual for a senior engineer from NVIDIA to attend an academic conference in Korea.
Similarly, Michelle Simmons, a professor at the University of New South Wales in Australia, presented before Mark Lam. Professor Simmons serves as the chief executive officer of Silicon Quantum Computing, a quantum computing corporation based in Australia. Representatives from the hottest semiconductor and quantum computing corporations in the tech field attended the AI academic conference.
Mark Lam highlighted AI as the key to NVIDIA's ability to showcase innovative GPUs. He said, 'NVIDIA is working to optimize the chip design process using AI,' adding, 'We are conducting research applying AI in various areas required for the chip design process, including physical design, logical design, concept validation, and packaging design.'
NVIDIA used AI to minimize voltage drop during the chip design process. Previously, it required setting up complex equations and repeatedly solving problems based on them, but NVIDIA reduced the time from three hours to three seconds using an AI neural network model. AI technology is also utilized in most processes, including predicting errors and optimizing designs.
Mark Lam emphasized, 'Just as GPUs enable AI, AI is also revolutionizing chip design, allowing for fast and precise chip designs that can dramatically enhance design productivity. AI will become increasingly important across semiconductor design in the future.'
Professor Simmons also stressed the importance of AI in developing quantum computers. As the company name suggests, Silicon Quantum Computing, which Professor Simmons founded, is creating quantum processing units (QPUs) based on silicon. She stated, 'In the past, silicon was considered a material that struggled to maintain quantum states, but through our research, we discovered that silicon can produce high-quality qubits and is favorable in the manufacturing process.'
Silicon Quantum Computing builds quantum computers in an ultra-high vacuum (UHV) environment. This process eliminates impurities, preventing contamination during manufacturing. AI also plays a crucial role in this process.
Professor Simmons explained, 'To build a quantum computer, we must consistently manufacture the same qubits and design a structure that can correct errors. After patterning one layer, we grow silicon and pattern again, stacking multiple layers. It is necessary to use AI to accurately predict the manufacturing process during this process.' Silicon Quantum Computing is continuously improving the accuracy of AI, which was initially around 90%.
Professor Simmons remarked, 'The silicon-based qubit technology we have developed operates with much higher accuracy than before and will become a core technology driving the future of quantum computing,' noting, 'We aim to launch our first commercial product in 2028 and develop a quantum processor capable of error correction by 2033.'