Key Takeaways
- Nvidia shares declined over 4% Thursday even as major tech companies announced dramatically higher AI infrastructure spending
- Google revealed intentions to offer its proprietary TPU chips to external buyers, intensifying competitive pressure
- Amazon highlighted rapid expansion of its internal chip development efforts
- Combined AI infrastructure investment from four leading hyperscalers could reach $725 billion in 2026
- Nvidia’s B300 server pricing in China has surged to approximately $1 million following intensified smuggling enforcement
Shares of Nvidia tumbled more than 4% Thursday, despite major technology companies unveiling unprecedented increases to their artificial intelligence infrastructure budgets. The decline signals mounting investor anxiety about Nvidia’s long-term competitive positioning as its largest clients accelerate development of proprietary chip solutions.
The selloff followed earnings releases from Meta, Alphabet, Microsoft, and Amazon, each announcing elevated capital expenditure forecasts for 2026. Meta increased its projection by $10 billion, establishing a new range between $125 billion and $145 billion. Alphabet elevated its guidance by $5 billion, reaching as high as $190 billion. Microsoft indicated that fourth-quarter capital spending alone would exceed $40 billion.
Collectively, these four hyperscale cloud providers now anticipate deploying up to $725 billion toward AI infrastructure throughout the year. With Nvidia controlling approximately 90% of the AI accelerator market, this spending surge would typically represent highly bullish news for the semiconductor manufacturer.
Yet investor sentiment doesn’t always align with fundamentally positive developments.
Alphabet’s TPU Strategy Triggers Market Concerns
The primary catalyst behind investor unease was Alphabet’s specific announcement regarding its Tensor Processing Unit technology. The company disclosed plans to market its custom TPU chips to select external enterprises, enabling installation within their private data centers.
Historically, TPUs served exclusively as internal infrastructure for Google. By commercializing these processors externally, Alphabet establishes them as a targeted competitor to Nvidia’s GPU offerings. While TPUs typically provide less versatility than Nvidia’s solutions, they deliver superior cost efficiency for specific artificial intelligence applications.
Amazon similarly emphasized accelerating momentum in its proprietary semiconductor initiatives during quarterly earnings discussions. Though both corporations remain significant Nvidia clients, the strategic trajectory appears unmistakable.
Nvidia has historically downplayed competitive threats from custom silicon, maintaining that its GPUs deliver superior adaptability for AI development workflows. However, sustaining that narrative without triggering market skepticism grows increasingly challenging.
Chinese Market Sees B300 Server Costs Approach $1 Million
Regarding supply dynamics, pricing for Nvidia’s cutting-edge B300 server systems within China has escalated to approximately 7 million yuan, climbing from roughly 4 million yuan late last year. This translates to nearly $1 million per configuration.
Intensified enforcement against semiconductor smuggling in China—which previously sustained a gray market for export-restricted hardware—has dramatically constrained availability. The B300 represents one of Nvidia’s most advanced AI server platforms and faces restrictions in China under United States export regulations.
Meanwhile, Thursday delivered mixed performance across the broader semiconductor industry. Qualcomm surged 9% following announcements of expanded data center initiatives. Memory manufacturers Sandisk, Western Digital, and Seagate also posted gains after Microsoft and Meta noted increasing expenditures for storage and memory infrastructure.
On the investment front, Nvidia’s venture capital division NVentures joined a $50 million extension of Swedish AI legal technology company Legora’s Series D funding round, establishing a $5.6 billion valuation and elevating total capital raised to $600 million.
Nvidia shares traded near $200.84 Thursday afternoon, representing a decline of approximately $8.41 for the session.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
