# AI%20infrastructure
Latest news and articles about AI%20infrastructure
Total: 52 articles found
Nvidia’s Jensen Huang Confirms Company Will Invest in OpenAI — but Not the $100bn Hype
Nvidia CEO Jensen Huang said the company will participate in the current OpenAI funding round and that the investment could be Nvidia’s largest ever, though it will be far below earlier $100 billion speculation. The pledge strengthens ties between a leading GPU supplier and a top AI-model developer, with strategic benefits and regulatory complexities for both firms and their customers.
AI Data‑centre Build‑outs Drive Chinese Optical‑Module Profits to Historic Peaks — But Growth Is Uneven
China's leading optical‑module makers have issued 2025 profit forecasts that reflect a surge in demand from AI data‑centre deployments. Xinyisheng and Zhongji Xuchuang expect historically large, core‑business‑driven profit gains, while Tianfu Communication posts solid but smaller growth, underscoring an uneven competitive landscape.
Musk Eyes ‘Space Compute’ and Robot Missions to Moon and Mars as He Weighs Corporate Consolidation
Elon Musk is reportedly advancing plans to shift substantial computing into space, consider mergers among his companies, and explore delivering Tesla’s Optimus humanoid robots to the Moon and Mars. The ambition combines SpaceX’s launch capability with Tesla’s robotics and Musk’s AI interests, but faces steep technical, economic and regulatory hurdles.
Alibaba Stakes a Claim in Nuclear Power to Secure an AI Advantage
Alibaba has taken a stake in a large nuclear power project as part of a broader push by Chinese tech firms to secure the baseload electricity needed for large AI deployments. China’s industrial capacity in power equipment and fast delivery gives its companies an edge in the global struggle for compute, reframing energy as a core element of AI competitiveness.
Shanghai Stakes a Claim to 120 ExaFLOPS of Smart Compute and Pitches Space–Ground 'Supernode' Architecture
Shanghai announced that its aggregate smart‑compute capacity has surpassed 120,000 petaFLOPS and outlined an agenda to build a self‑reliant compute ecosystem linking domestic chips, models and cloud services. Officials and industry groups also floated an integrated space–ground architecture that would treat LEO satellites as programmable compute clusters to achieve global, low‑latency coverage for selected applications.
AI’s Hunger for Memory Could Keep Global Chip Shortages Dragging On Until 2027
Synopsys CEO Sassine Ghazi warns that the current memory-chip shortage, driven by heavy demand from AI data centres, is likely to last through 2026 and potentially into 2027. Concentrated production, long lead times for new fabs and booming demand for HBM mean elevated prices and allocation pressures may persist, benefiting memory suppliers but squeezing device makers and other industries.
Samsung Doubles NAND Prices as AI‑Fueled Storage Supercycle Tightens Supply
Samsung has raised NAND flash prices by over 100% in Q1 2026 as AI‑driven demand for high‑performance storage outstrips supply. Analysts say the industry has entered a storage‑chip "supercycle," with tight capacity likely to persist until at least 2027 and meaningful new supply not expected until 2028.
From Shanghai Markets to Launchpads: Jensen Huang’s China Stopover and a Boost for Beijing’s Commercial Space Push
Jensen Huang’s early-2026 visit to Shanghai and public-facing interactions underline ongoing commercial links between Nvidia and China despite geopolitical headwinds. At the same time, Beijing is accelerating support for a domestic commercial space sector — highlighted by policy measures for satellite-data use and Zhongke Yuhang’s completion of IPO counselling — while tightening data and security regulation, creating both opportunities and risks for firms operating in China.
vLLM Founders’ New Startup Raises $150m Seed at an $800m Valuation — A Big Bet on LLM Infrastructure
Inferact, founded by the vLLM core team, raised $150 million in a seed round led by Andreessen Horowitz and Lightspeed at an $800 million valuation. The deal highlights investor enthusiasm for LLM inference and deployment infrastructure, but sets high expectations for rapid commercialisation amid fierce competition and regulatory questions.
NVIDIA’s Jensen Huang Says AI Build‑out Could Lift Tradespeople Into Six‑Figure Pay Brackets
NVIDIA CEO Jensen Huang warned that the build‑out of AI infrastructure will push demand for on‑site skilled trades—electricians, plumbers and HVAC technicians—so high that they could earn six‑figure annual salaries. The observation underscores that AI’s expansion creates major labour and logistical pressures in the physical infrastructure layer, with implications for wages, training and the costs of deploying large models.
Nvidia’s Jensen Huang at Davos: AI Is Not a Bubble—It’s a Trillion‑Dollar Infrastructure Build
At Davos, Nvidia CEO Jensen Huang argued that current AI spending is the start of a vast infrastructure build rather than a speculative bubble, outlining a five‑layer model from energy to applications. He predicted trillions in additional investment, flagged GPU shortages and supply‑chain pressures, and pushed for national "AI sovereignty" while saying automation will create high‑paid technical roles even as some white‑collar jobs are displaced.
China Stocks Open Higher as Chip and Storage Names Lead a Tech-Heavy Rally
Chinese stock markets opened higher as semiconductor and storage‑chip stocks outperformed, buoyed by company profit forecasts and plans to expand wafer‑level packaging capacity. Broader market moves were uneven, with commodity and precious‑metals names lagging and speculative small‑caps showing continued volatility.