Nvidia CEO Jensen Huang Reveals New AI Chip Blackwell GPU Priced at $30,000-40,000 for Wider Customer Base
In an exclusive interview on Tuesday, Nvidia Corp. CEO Jensen Huang shared details about the company’s latest generation of AI training chip, the Blackwell GPU. Huang expressed optimism about the growth of artificial intelligence (AI) and how the new chip will attract a wider customer base at a reasonable price.
The Blackwell GPU is NVIDIA’s flagship product following the Hopper architecture data center GPU released earlier this year. It is expected to replace the market position of the H100 and is named “B200” skipping the 100 sequence.
Huang stated that the B200 may be priced between $30,000 and $40,000, which is lower than some analysts expected. He emphasized that the pricing was set to attract as many customers as possible but noted that prices will change based on the value provided.
The CEO also mentioned that Nvidia is working with TSMC to avoid chip packaging bottlenecks, ensuring smoother shipments for the new chips. He anticipates the B200 to be available later this year and assured analysts of a good understanding of the demand for Blackwell chips.
Nvidia’s Chief Financial Officer, Colette Kress, added that the new chip and its system will be available later this year, acknowledging potential supply constraints despite better preparation for increased demand.
Huang also discussed Nvidia’s growth in the global data center market, expecting a high share of revenue from global investment in data center equipment. He highlighted the company’s focus on developing software to enable AI technology adoption across various industries, anticipating a significant business opportunity in this area.
In conclusion, Huang emphasized Nvidia’s transition from selling individual chips to entire systems, aiming to capture a larger portion of the data center market. He stated, “Nvidia doesn’t make chips, it builds data centers.”
(Source: Financial Associated Press)