All articles tagged with #ai data centers
A sweeping US heatwave strains the power grid and water supplies as the nation races to build thousands of AI data centers, whose cooling and electricity needs could lift overall energy demand and stress local water resources, prompting calls for greater oversight, siting checks, and faster upgrades to generation and transmission, as utilities implement measures like backup power dispatch.
Samsung Electronics and SK Hynix tumbled on a broad Nasdaq-driven tech selloff, wiping out billions in value and pulling Korea’s Kospi lower. SK Hynix outlined a large Korea-focused investment plan to expand memory and AI chip production (M17 and related facilities) as it and the industry bet on rising demand for AI infrastructure, with government backing a nationwide push toward AI data centers and a national semiconductor ecosystem; SK Hynix is also preparing to begin trading ADRs on Nasdaq.
A major US heatwave is expected to push electricity use to near-record levels this week, straining grids operated by PJM Interconnection, NYISO, and MISO. Officials warn that demand is rising faster than new supply, a gap amplified by hyperscale AI data centers—concentrated in northern Virginia—that increase energy use and contribute to a data heat island effect. Health advisories accompany the forecast, with peak demand around 166.3 gigawatts projected for Thursday evening as temperatures climb above 38 C across much of the eastern half of the country.
Progressive NYC candidates endorsed by Mayor Mamdani swept key House primaries, backing a wealth-tax approach for billionaires, a moratorium on AI data centers, a 32-hour four-day workweek, and labor-law reforms through the PRO Act to boost unions and curb wealth inequality.
South Korea unveiled a $1 trillion push combining expanded memory-chip production, new AI data centers, and commercial humanoid robots by 2028, backed by Samsung, SK Hynix, Hyundai and others. The plan includes about $585 billion for new chip fabs to double DRAM and $357 billion for large AI data centers, plus Hyundai’s $5.8 billion investment to scale Atlas robots to up to 30,000 per year. Authorities also seek a national physical-AI program and a Korean foundation model, but face power/water needs and labor concerns over robotics impacting jobs.
Samsung Electronics and SK Hynix are reportedly set to unveil up to 2,000 trillion won ($1.3 trillion) in 10-year investment plans, including new fabs and AI data centers, triggering stock declines; the announcement will be made at a government briefing chaired by President Lee Jae Myung and follows earlier reports of a 1,000 trillion won Samsung plan, underscoring Korea’s push to expand AI chip capacity.
IDC expects AI data centers to absorb about 70% of global memory-chip production in 2026, squeezing supplies for phones, PCs, and other devices. The two primary U.S.-listed plays are Micron Technology (MU) and SK Hynix (A000660), which dominate high-bandwidth memory (HBM4) used by Nvidia and hyperscalers; Micron reports its HBM production is sold out under multiyear contracts and has seen DRAM/NAND price increases, while SK Hynix commands a large share of HBM4 volume and is co-developing next-gen memory with Nvidia. Bank of America sees robust growth for the HBM market (potentially $54.6B in 2026). SK Hynix is preparing to list ADRs in the Nasdaq, adding accessibility for U.S. investors. The piece frames MU and SK Hynix as the standout long-term AI infrastructure plays, with a suggested dollar-cost averaging approach, though it notes Motley Fool Stock Advisor did not currently name MU among its top picks.
Microsoft is sliding in June despite solid revenue and earnings as capital spending on AI data centers squeezes margins and free cash flow; hyperscalers are projected to spend hundreds of billions in 2026, shifting investor focus from current earnings to the future costs of AI infrastructure.
Federal regulators are backing the Trump administration’s plan to accelerate electric service and grid interconnections for energy-hungry AI data centers, aiming to meet surging demand and speed infrastructure, while signaling faster permitting—though the move raises questions about grid reliability and environmental reviews.
A decade of diplomacy, including a Trump-era push with Paraguay and plans for U.S.-backed AI data-center projects, appears to be steering North America’s 2026 World Cup, while side stories about Bosnian Utica and protests in Mexico illustrate how politics and sport intertwine in this global event.
OpenAI researchers say China-backed operatives used ChatGPT to generate social-media content aiming to influence U.S. perceptions of AI data centers and related policy, with two clusters of China-originated accounts promoting narratives that data centers raise electricity costs and that tariffs affect tech leadership. Posts were posted by batches of accounts posing as Americans and have since been banned. While the effort reflects China’s attempt to sway the domestic debate, OpenAI found no evidence it significantly shifted public opinion. The findings feed into ongoing U.S. political and regulatory discussions about AI infrastructure, energy impacts, and national tech competition.
Goldman Sachs says private infrastructure and real estate will play a larger role in financing the AI-driven data-center boom, lifting capex forecasts for Meta, Microsoft, Amazon and Alphabet to about $5.3 trillion for 2025–2030 (up from $4.5 trillion). Funding will come from public, securitized and private markets as data-center projects blur lines across land, power, buildings and equipment. Private infra’s inflation-protected income could accelerate growth, with assets under management expected to top $3 trillion by 2030, after an 11.5% annualized rise 2021–2024.
NVIDIA CEO Jensen Huang used Marvell’s Computex keynote to champion a copper-to-optics strategy for AI data centers—use copper where possible, then scale with optics to meet GPU-to-GPU traffic needs. The comments helped lift optical-interconnect names (Coherent, Lumentum, Corning, Marvell, Ciena) as investors expect a rapid shift to photonics in AI infrastructure, backed by Nvidia’s multi‑billion‑dollar photonics investments and bullish Bank of America price targets for related stocks.
The Electric reports that rising AI data-center workloads are creating a new revenue stream for Western battery makers by deploying backup power systems for AI facilities, potentially triggering a broader battery boom even as consumer EV demand remains tepid.
Micron Technology's stock climbs after Samsung reaches a last-minute labor deal averting an 18-day strike that could disrupt memory-chip supply. A surge in demand for high-bandwidth memory for AI data centers helps the sector, with Micron reporting record Q2 revenue of $23.9 billion, EPS $12.07, and near doubling of gross margins as new plants come online. Wall Street shows a Strong Buy consensus, with an average target of about $657 per share, implying some downside from current levels.