Nvidia to expand capacity through next year: CEO

Nvidia Corp CEO Jensen Huang revealed plans to expand production capacity by twofold in the second half of 2024, with even greater increases expected for the second half of 2025. Speaking at Taipei’s Songshan Airport before departing Taiwan after a 14-day visit, Huang emphasized that supply chain constraints stem from unprecedented demand growth. He noted that Nvidia’s AI computing Grace Blackwell platform is now in commercial production, while mass production of the next-generation Vera Rubin platform has begun. Huang arrived in Taipei on May 23 to meet with key suppliers, including Taiwan Semiconductor Manufacturing Co (TSMC) and AI server maker Hon Hai Precision Industry Co. The visit coincided with Nvidia GTC Taipei and Computex Taipei, where he praised Taiwan’s critical role in advancing AI technology. ‘Capacity is tight all over the supply chain because demand is so strong and we’re growing so fast,’ Huang stated, adding that the second half of 2025 will see massive production demands. The CEO highlighted Nvidia’s reliance on Taiwan’s semiconductor ecosystem, calling it essential for the trillion-dollar AI computing industry. He also acknowledged strong partnerships in South Korea, where he traveled after Taiwan, his second visit in seven months. Memory capacity shortages across the supply chain remain a challenge, Huang said, underscoring the need for global preparedness. During his trip, Huang attended Computex Taipei, where Nvidia showcased its latest AI advancements, including the GB300 Grace Blackwell Ultra Superchip featuring SK Hynix’s HBM3E memory. The event reinforced Taiwan’s position as a hub for AI innovation, with Huang expressing gratitude for the country’s support in driving the next computing revolution.