Nvidia stock was volatile on Thursday after Chief Executive Jensen Huang cautioned that Chinese chipmakers are rapidly catching up to US firms in artificial intelligence capabilities, a warning that rattled investor confidence in the AI trade.
The stock opened slightly higher, up 0.5% in early trading, before losing ground following a 2% decline on Wednesday.
Broader AI-linked equities were mixed, as investors reassessed how long current valuations can hold given rising competition and slowing hardware demand.
Nvidia remains a primary beneficiary of AI infrastructure spending, but the sector’s volatility has started to weigh on investor sentiment.
The company continues to trade near record highs despite heightened scrutiny over its growth prospects and restrictions on global chip exports.
Huang: “China is nanoseconds behind America in AI”
Huang’s remarks, published late Wednesday, struck a nationalistic tone, urging stronger domestic support for the US AI industry.
“As I have long said, China is nanoseconds behind America in AI,” Huang wrote on X (formerly Twitter).
“It’s vital that America wins by racing ahead and winning developers worldwide.”
The statement was widely interpreted as a veiled call for fewer export restrictions on Nvidia’s advanced AI chips.
The company remains barred from selling its top-end processors to China under US trade rules, limiting a key source of revenue growth.
The issue extends beyond China. While Microsoft recently secured approval to export Nvidia’s GB300 chips to the United Arab Emirates, Washington continues to cap the total computing power allowed in shipments to certain regions, including parts of the Middle East.
Amid those limits, Huang’s comments were seen as a rallying cry for maintaining America’s technological lead — and a warning that tightening restrictions could backfire by enabling competitors abroad.
Google ups the ante with new AI chip
Adding to the competitive pressure, Google on Thursday announced that it will make its most advanced AI chip publicly available, expanding its challenge to Nvidia’s dominance in high-performance computing.
The seventh-generation Tensor Processing Unit (TPU), known as Ironwood, will be released for public use in the coming weeks, the company said.
Initially unveiled in April for internal testing, Ironwood is designed to handle demanding AI workloads ranging from model training to real-time chatbot operations.
Google said the new chip can connect up to 9,216 TPUs in a single pod, eliminating “data bottlenecks for the most demanding models” and allowing customers to “run and scale the largest, most data-intensive models in existence.”
The company claims Ironwood is more than four times faster than its predecessor and has already secured major customers, including AI startup Anthropic, which plans to use up to 1 million Ironwood TPUs to train its Claude model.
Google’s move underscores the intensifying race among US tech giants to dominate AI infrastructure.
The company is competing directly with Microsoft, Amazon, and Meta Platforms to supply the computing backbone for the next generation of large language models and autonomous systems.
While most AI workloads today rely on Nvidia’s graphics processing units (GPUs), custom chips like Google’s TPUs offer potential advantages in cost, efficiency, and scalability.
Alongside Ironwood, Google announced a suite of upgrades to its cloud platform, promising faster, cheaper, and more flexible AI deployment options — a direct bid to win clients from Amazon Web Services and Microsoft Azure.
The post Nvidia stock is sliding over 2% today: what’s hurting the AI darling today? appeared first on Invezz