Google Gemini Co-Lead Noam Shazeer Leaves for OpenAI Amid AI Talent Race 
Meta Seeks Legal Shield From Child-Harm Lawsuits Amid KOSA Talks 
SK Hynix Shares Hit Record High After Shipping Next-Generation HBM4E AI Memory Samples 
Obayashi to Acquire Multiplex in $526M Expansion Deal 
SpaceX Surpasses Amazon in Market Value as Post-IPO Rally Accelerates 
BHP Shares Fall as Jansen Potash Project Costs Surge 
Kingboard Holdings Shares Surge After HK$11.77 Billion Block Trade to Expand PCB and AI Supply Chain Business 
Apple Signals Product Price Hikes Amid Rising Memory Chip Costs 
SpaceX Stock Slides After IPO Rally as Valuation Concerns Grow 
SoftBank Shares Drop as OpenAI Losses and Rising Costs Spark Investor Concerns 
US Raises Concerns Over Possible ASML EUV Machine Transfer to China 
Anthropic Officials Meet White House Over AI Model Outage 
Saudi Aramco Explores Sulphur Business Stake Sale to Raise Billions 
Trump Administration Delays DeepSeek and CXMT Trade Blacklist Designations Amid U.S.-China Tensions 
Qantas Unveils Wellness-Focused Nonstop Sydney-London Flights to Reduce Jet Lag 
OpenAI's $34B Spending Pushes AI Market Leadership Ahead of IPO 
Chinese Social Media Giant Xiaohongshu Eyes Hong Kong IPO at Over $70 Billion Valuation 
Nvidia is reportedly developing a version of its Groq-based artificial intelligence chips tailored for the Chinese market, according to two sources familiar with the matter who spoke to Reuters. The move signals a significant strategic push by the chipmaking giant to expand its footprint in one of the world's largest technology markets.
The development follows Nvidia's licensing agreement with AI chip startup Groq, finalized late last year in a landmark $17 billion deal. At its annual developer conference held this week in San Jose, California, Nvidia unveiled a new lineup of products built around Groq's chip architecture, primarily targeting the inference segment of the AI market — where systems respond to queries, generate code, and execute user-driven tasks.
Nvidia CEO Jensen Huang confirmed during his keynote address that the company has resumed production of its H200 chips — the predecessor to its current flagship — after securing export licenses from the Trump administration and locking in purchase orders from Chinese buyers. "The inference inflection has arrived," Huang stated, projecting the revenue potential for Nvidia's AI chips could surpass $1 trillion by 2027.
One source clarified that the chips being prepared for China are not downgraded or market-specific versions. Instead, they are adaptable variants designed to integrate with a broader range of systems, with availability expected as early as May. These chips will be paired alongside Nvidia's upcoming Vera Rubin processors in its domestic lineup — hardware that currently remains restricted from Chinese export.
The inference chip market presents growing competition for Nvidia, with major Chinese technology companies like Baidu already producing proprietary inference solutions. By leveraging Groq's technology, Nvidia aims to establish a stronger competitive position in a space where its dominance is less entrenched compared to AI model training. Nvidia has not yet issued an official comment on the matter.
