TikTok’s Chinese internet technology parent company ByteDance plans to spend $7 billion this year on Nvidia’s advanced artificial intelligence (AI) chips.
If secured, ByteDance will become one of the world's largest owners of Nvidia chips.
The news was announced this week despite U.S. export restrictions, which limit China’s access to the technology.
Around the same time, Nvidia had just completed its $700 million acquisition of Run.ai, an AI workload and resource management platform.
We’re delighted to welcome the Run:ai team to NVIDIA. https://t.co/us7p6rnvYl
— NVIDIA (@nvidia) December 30, 2024
With the AI chip acquisition, ByteDance hopes to build its own AI models — much like its Doubao chatbot app which boasts about 60 million monthly users.
In addition, the Nasdaq reports that ByteDance plans to scale its operations internationally, as its founder, Zhang Yiming has been negotiating with suppliers and data center operators in Southeast Asia and Europe to bolster its infrastructure.
In total, the company plans to invest more than $20 billion in AI-related hardware this year, which includes chips, data centers, and undersea cables.
Driving Forward with AI
If successful, ByteDance could position itself to compete with current industry giants in the AI space. The strategic move would allow it to expand its operations and infrastructure.
By securing chips through overseas cloud services, the company could also bypass export restrictions by running its data centers outside of China.
At the start of 2024, ByteDance announced plans to grow its business tenfold to $17.5 billion, putting TikTok in direct competition with Amazon, Shein, and Temu in the American market.
However, the future of TikTok's U.S. penetration hangs in the balance as President Donald Trump requested to delay the app's impending countrywide ban, which was scheduled for January 19, 2025.
Google’s recent unveiling of Gemini 2.0 is another demonstration of the demand for advanced technologies in the AI sector.
Gemini 2.0, its most advanced AI model to date, is designed to push the boundaries of natural language understanding, multimodal capabilities (working across text, images, and video), and real-time problem-solving.