TikTok Parent ByteDance Reportedly Invests $2.5B in Nvidia B200 AI Expansion
TikTok's parent company is expanding its AI training infrastructure by deploying thousands of high-performance Nvidia B200 GPUs at a facility in Malaysia.
ByteDance, the Beijing-based parent company of TikTok, is expanding its AI compute infrastructure with access to Nvidia’s most advanced AI GPUs.
According to a Wall Street Journal report published Thursday, the company is deploying Nvidia Blackwell chips at a data center in Malaysia, marking a significant step in its push to scale AI model development.
Increased AI compute power is expected to help ByteDance accelerate model training and improve services across its digital ecosystem, including content technologies and algorithmic recommendations used by TikTok.
ByteDance Nvidia Chips Access
ByteDance is working with Southeast Asian firm Aolani Cloud to deploy about 500 Nvidia Blackwell computing systems in Malaysia, totalling roughly 36,000 B200 chips and related hardware for machine-learning workloads, according to the Wall Street Journal, citing people familiar with the matter.
The hardware is acquired through Aivres, a company that assembles servers using Nvidia chips. If the plan is fully implemented, the hardware involved could cost more than $2.5 billion. Aolani currently operates with about $100 million in hardware, a spokesman told the WSJ.
According to Yahoo Finance, ByteDance plans to use the computing power for AI research and development outside China and to meet growing global demand for AI among its customers.
The development follows earlier reporting that ByteDance had also been using B200 Nvidia chips at a data center in Indonesia, suggesting a deliberate strategy of building AI compute capacity across Southeast Asia.
For ByteDance, the infrastructure upgrade supports research, training pipelines, and AI-driven features used across its platforms.
The Significance of ByteDance’s $2.5 Billion Bet
The Malaysia deployment signals that ByteDance is aggressively scaling its AI training hardware capacity at a time when demand for high-performance GPUs is intensifying across the global technology industry.
The Nvidia B200, part of the Blackwell architecture, represents the current frontier in AI data center infrastructure, purpose-built for large-scale model training and inference workloads.
As noted by Investing.com, the TikTok owner is ramping up its AI offerings with the goal of competing in the industry at a global scale, having released several AI apps in China and overseas, with its AI video generation model, Seedance, recently gaining viral attention.Securing access to top-tier AI compute power is central to sustaining that ambition.
The scale of this deal, potentially exceeding $2.5 billion in hardware, underscores how seriously ByteDance is treating AI infrastructure as a competitive foundation.
Executive Voices and Market Reaction
Neither ByteDance nor Nvidia issued public statements in response to the Wall Street Journal’s reporting. Reuters noted it could not immediately verify the report, and Nvidia, ByteDance, and Aolani Cloud did not immediately respond to its request for comment.
Industry analysts referenced in financial coverage from Investing.com described access to Nvidia’s latest GPUs as an important step for companies seeking to scale advanced AI systems
Separate investment signals also point in that direction. According to the South China Morning Post, ByteDance previously indicated plans to spend about 100 billion yuan, roughly $14 billion, on advanced AI processors in 2026, up from around 85 billion yuan in 2025.
Technology market observers note that demand for high-performance GPUs remains extremely strong as more companies build internal AI platforms. Nvidia’s hardware is widely considered the industry standard for large-scale AI training, giving companies with access to these systems a strategic advantage.
Broader Impact of ByteDance Investment
The deployment of tens of thousands of Nvidia B200 chips through a third-party operator in Malaysia has clear implications for how global AI infrastructure is being built and distributed.
For developers and enterprise customers building on ByteDance’s AI platforms, expanded compute capacity translates directly to faster model iterations and more reliable service delivery.
As noted by Tom’s Hardware Chinese companies have increasingly attempted to build AI data centers outside their home country to access hardware unavailable domestically, and ByteDance’s Malaysia arrangement reflects that pattern.
For the broader AI ecosystem, the deal also highlights the central role that GPU supply chains and the companies that assemble and distribute Nvidia hardware play in determining who can compete at the frontier of AI development.
What’s Next For ByteDance
ByteDance’s AI infrastructure buildout shows no signs of slowing. The company is expected to continue expanding data center capacity across Southeast Asia as it scales its AI product lineup globally.
Regulatory and supply chain developments will remain closely watched, particularly any formal agreements between Nvidia and ByteDance regarding chip procurement conditions.
NVIDIA, ByteDance, and Aolani Cloud have yet to publicly comment on the reported Malaysia deal, and official confirmation could bring additional details about the scope and timeline of the deployment.
![Top Tech Stories of 6th week [2026]](https://www.nogentech.org/wp-content/uploads/2026/02/Top-Tech-Stories-of-6th-Week-2026-390x220.webp)