Nvidia CEO Backs Samsung Despite AI Memory Chip Setbacks
Sign up for ARPU: Stay ahead of the curve on tech business trends.
Nvidia CEO Jensen Huang expressed confidence in Samsung's ability to overcome challenges in producing high-bandwidth memory (HBM) chips for AI systems, despite recent production difficulties, reports Bloomberg. While acknowledging Samsung's struggles to meet Nvidia's standards for HBM, Huang expressed confidence in the company's capacity to innovate.
"They have to engineer a new design," Huang stated at a press briefing at CES in Las Vegas. "But they can do it. They are working very fast. They're very committed to do it."
Samsung's slower progress in HBM production has impacted its performance in the lucrative AI market. The company's latest quarterly results, released Wednesday, fell short of analysts' estimates. This underperformance highlights the significance of HBM in today's AI systems.
AI computing systems powered by Nvidia have witnessed tremendous growth in the past two years. These systems rely on Nvidia chips as their central processing units, utilizing vast amounts of data to train AI models. Memory chips play a critical role in supplying the processors with the necessary information.
The drive to increase memory speed and capacity, and to integrate components more closely with processors, has introduced significant complexity in chip production. Samsung has been working to overcome these obstacles.
"They're going to succeed – no question," Huang said. "I have confidence that Samsung will succeed with HBM."
Samsung, while the world's largest overall memory chip manufacturer, has faced challenges in catching up to competitors like SK Hynix in the AI market. This highlights the importance of HBM in AI systems, where Nvidia chips currently dominate.