TAIPEI (Taiwan News) — Nvidia founder and CEO Jensen Huang (黃仁勳) dismissed concerns over competition from custom silicon chips during a dinner with top Taiwanese technology executives in Taipei on Saturday, saying such products pose little threat to Nvidia’s dominance in AI computing, per UDN.

Huang spoke with local media after the gathering, which was dubbed a “trillion-dollar dinner” because of the combined net worth of attendees. Guests included TSMC Chair and President C.C. Wei (魏哲家), whose company is Nvidia’s longtime manufacturing partner.

Custom silicon, also known as application-specific integrated circuits, is increasingly used by major cloud service providers such as Google, Meta, and Amazon for specific workloads. While these chips are cheaper and more energy-efficient than general-purpose graphics processing units, Huang said Nvidia’s advantage lies in building the entire AI infrastructure, not just GPUs.

“Nvidia is collaborating with all leading AI companies,” Huang said, naming Google, OpenAI, and xAI. He added that it was irrelevant whether some cloud providers are also considered Nvidia competitors.

Huang said Nvidia technology is now ubiquitous, used across cloud platforms, computer systems, robots, and automobiles worldwide. He emphasized the company’s rapid pace of innovation and said Nvidia’s technology offers the lowest overall cost.

He added that Nvidia’s cumulative shipments of GPUs are unlikely to be surpassed by ASICs, citing the company’s scale and investment in research and development. Nvidia’s annual R&D budget is about US$20 billion (NT$632 billion) and could grow by as much as 50% per year, Huang said.

Huang also highlighted Nvidia’s software ecosystem, saying all AI developers rely on Nvidia platforms. He said the company supports an estimated five million to six million developers globally and will continue to expand that support.