GTC felt more bullish than ever, but Nvidia’s challenges are piling up

Nvidia made a significant impact in San Jose this year, attracting a record-breaking 25,000 attendees to the San Jose Convention Center and surrounding downtown buildings. The event featured packed workshops, talks, and panels, with attendees even having to lean against walls or sit on the floor due to the high turnout.

Currently dominating the AI industry, Nvidia boasts impressive financials, high profit margins, and a lack of serious competitors. However, the company faces challenges ahead, including U.S. tariffs, competition from DeepSeek, and changing priorities among top AI customers.

At GTC 2025, Nvidia CEO Jensen Huang showcased confidence by unveiling new powerful chips, personal “supercomputers,” and adorable robots in a comprehensive sales pitch aimed at investors amid a decline in Nvidia’s stock.

Inference boom

Nvidia emphasized at this year’s GTC that the demand for its chips will continue to rise. Huang addressed concerns about traditional AI scaling becoming outdated, highlighting DeepSeek’s efficient “reasoning” model but asserting that power-hungry models will increase the need for Nvidia chips.

Despite the competition from companies like Cerebras and Groq in the low-cost inference hardware market, Nvidia believes its new line of Vera Rubin GPUs will excel in running AI models, maintaining its position in the industry.

However, Nvidia faces challenges from emerging competitors and the efforts of tech giants like OpenAI and Meta to develop their custom chips for AI applications, potentially weakening Nvidia’s market dominance.

Tariff tensions

Addressing concerns about tariffs, Nvidia reassured attendees at GTC 2025 that the company is prepared for any potential impact. Despite the U.S. not imposing tariffs on Taiwan, where Nvidia sources most of its chips, uncertainties remain about the long-term economic effects.

Pledging to invest in U.S. manufacturing, Nvidia aims to diversify its supply chains but faces the challenge of maintaining healthy profit margins given the significant costs involved.

New business

In an effort to expand beyond its core chips business, Nvidia highlighted its investments in quantum computing at GTC, acknowledging the importance of the industry. The company also introduced new personal AI supercomputers like DGX Spark and DGX Station, targeting users for AI model prototyping and development.

Additionally, Nvidia announced the establishment of a new quantum computing center in Boston, NVAQC, to collaborate with industry leaders and advance quantum computing using Nvidia chips.

Nvidia envisions personal AI supercomputers as a potential revenue stream, with products like DGX Spark and DGX Station offering AI model capabilities at the edge, though at a premium price point. Huang sees these products as the future of personal computing, signaling a shift in the industry.

The response from customers to Nvidia’s new initiatives will be crucial in determining the company’s future success.