NVIDIA Stock Price Soars 8%
Jensen Huang, CEO of Nvidia, the leading company in artificial intelligence (AI), mentioned the possibility of entrusting Samsung Electronics' foundry (semiconductor contract manufacturing) with the production of AI chips.
On the 11th (local time), Huang gave a keynote speech at the Goldman Sachs Group-hosted Tech Conference held at the Palace Hotel in San Francisco, California, stating that Nvidia's most popular 'Hopper' series (H100·H200) and the next-generation chip 'Blackwell' are all currently entrusted to the world's largest foundry, Taiwan's TSMC.
Regarding the reliance on TSMC for the production of these chips, Huang explained, "Because TSMC is the best in the industry." He added, "TSMC's agility and ability to respond to our demands are amazing."
However, he said, "If necessary, we can use other companies at any time." Although Huang did not specifically mention other companies, currently, the only companies capable of producing Nvidia's latest chips besides TSMC are Samsung Electronics. Therefore, it is interpreted that Huang hinted at the possibility of entrusting Samsung Electronics with AI chip production.
Bloomberg interpreted Huang's remarks in connection with the potential geopolitical risks in Taiwan. Concerns are spreading that China might unify Taiwan by force, and if this materializes, Nvidia could face disruptions in AI chip production. However, Huang also added, "Such a change (switching the manufacturer of Nvidia AI chips) could potentially lead to a decline in chip quality."
Huang also emphasized the strong demand for the latest chip Blackwell, which aims for mass production within this year. He expects to generate billions of dollars in revenue from Blackwell in the fourth quarter of fiscal year 2025 (from November 2024 to January 2025).
Regarding the recent AI bubble theory, which arose from concerns that big tech's AI investments are excessive compared to demand, he dismissed it by saying, "Companies have no choice but to adopt 'accelerated computing' (a technique that significantly improves processing speed through specialized hardware)." He added, "Our (Nvidia's) technology not only accelerates existing data processing but also handles AI tasks that existing technologies cannot."
On the same day, Nvidia's stock price on the New York Stock Exchange closed at $116.78, up 8.03% from the previous day, marking the largest single-day gain since July 31. The Philadelphia Semiconductor Index, a global semiconductor index, rose 4.90%, but TSMC's stock price showed a slight decline, seemingly affected negatively by Huang's remarks.
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
