If you want an example of “the right product at the right time”, Nvidia’s artificial intelligence-focused H100 GPUs can be shown. While the company has achieved incredible success with its artificial intelligence products, it intends to continue this in the future. According to the statements, Nvidia plans to sell 1.5 million to 2 million H100 units by 2024.Right time, right product, record profit
Just a few hours ago, we told you that the sales of Nvidia’s data center division have increased by more than 170 percent. While the company’s second quarter results exceeded all expectations and rose to a record level, a similar performance is expected in the current quarter.
Nvidia CEO Jensen Huang touched on this area in his statement and said: “Our demand is enormous. We are expanding our production capacity significantly. Supply will increase significantly for the rest of this year and next year. Nvidia has been preparing for this for over two decades and has created a new computing platform on which the world industry can build.”
According to the Financial Times report, Nvidia is working to significantly upgrade its manufacturing facilities to produce high-volume AI GPUs next year. To give an idea of this increase, Nvidia will ship 550,000 H100 GPUs this year, and the company plans to nearly triple that amount next year. And already until December, the order calendar is full.The explosion in artificial intelligence has put all companies in the race for productive AI, and Nvidia has been the company to make the most of it so far. In a situation where AI chip orders are booked through 2024, the real question is how Nvidia will fulfill such large orders, given that the company is facing several challenges.
Samsung’s growing importance
Nvidia is currently a partner of TSMC and is responsible for the production of the semiconductor giant green team’s artificial intelligence GPUs. The problem is that TSMC doesn’t just work with Nvidia, it also takes orders from Apple and AMD. Although TSMC plans a rapid expansion to meet the needs of the industry, this will not materialize until 2024. So Nvidia has to adopt another plan and shift to dual-source production.
We reported yesterday that Samsung is in talks with AMD to take orders for the Instinct MI300X AI accelerators. Unlike TSMC, which outsources components such as HBM, Samsung has become the focus of attention in the industry as it offers a “hybrid” approach to its customers, taking responsibility for all development phases. Therefore, Nvidia can turn to Samsung to achieve its sales targets.
29636
