Despite the latest financial report showing that Nvidia's performance guidance for the third quarter did not meet investors' expectations, the stock price plummeted nearly 7% after hours. But a senior executive of the company remained optimistic and described new clients from around the world, including governments seeking to develop their own artificial intelligence (AI) models.
Nvidia expects its third quarter revenue to be approximately $32.5 billion, which will increase by 80% compared to the same period last year and exceed the average analyst estimate of $31.7 billion. However, the market's highest expected revenue for Nvidia in the third quarter reached $37.9 billion, which has raised concerns that its explosive growth is weakening.
After the release of the financial report, Colette Kress, Chief Financial Officer of Nvidia, stated in a conference call with analysts that as generative artificial intelligence becomes the center stage for businesses and governments, countries are building AI models in their own languages, further accelerating the already booming demand.
She explained that in the fiscal year ending January 2025, countries adopting their own artificial intelligence applications and models will contribute approximately tens of billions of dollars (in the low range) in revenue to Nvidia. The previous prediction was that such sales would contribute up to several billion (high range) dollars to total revenue.
Each country in the world hopes to have its own generative artificial intelligence that can integrate its own language, culture, and data into its own country, "Kress added, describing AI expertise and infrastructure as a" national priority.
She cited the example of the National Institute of Advanced Industrial Science and Technology (AIST) in Japan, which is building an artificial intelligence supercomputer equipped with thousands of Nvidia H200 graphics processors.
At the same time, governments around the world also consider artificial intelligence as a measure to enhance national security.
IDC Computing Semiconductor Analyst Shane Rau said, "Artificial intelligence models are trained based on data, and for political entities, especially countries, their data is confidential. Their models need to be customized according to their unique political, economic, cultural, and scientific needs
Therefore, they need to have their own artificial intelligence models and customized underlying hardware and software arrangements, "he said.
TECHnalysis Research Chief Analyst Bob O'Donnell stated that in addition to large cloud providers such as Microsoft investing heavily in Nvidia hardware, countries hoping to create their own artificial intelligence models can bring growth opportunities for Nvidia's GPUs.
Finally, companies have been striving to leverage government initiatives to establish artificial intelligence platforms in regional languages. In May of this year, IBM announced that Saudi Arabia's Data and Artificial Intelligence Authority would use the company's artificial intelligence platform Watsonx to train its "ALLaM" Arabic model.