Gigawatt campuses, power grids, cooling and the trillions being poured into AI compute buildouts.
516 articles
Beijing has officially established a new AI factory infrastructure project. The facility aims to achieve 100,000 P-Flops of computing power and produce 10 trillion tokens daily.
量子位 · 94d ago
Intel introduced the 18A-P process node, an enhancement of 18A, offering significant power and performance gains. The process will debut in the 'Diamond Rapids' product line.
cnBeta (Full RSS) · 94d ago
Hewlett Packard Enterprise (HPE) has launched new networking hardware specifically designed for AI data centers. The company is leveraging technology from its Juniper Networks acquisition to capture growing corporate demand for AI infrastructure.
Bloomberg Technology · 95d ago
The UK energy regulator is considering mandates for data centers to reduce power consumption during peak stress periods. This move addresses the massive electricity demand driven by AI and cloud computing growth.
Bloomberg Technology · 95d ago
Nvidia is launching a massive $25 billion debt sale to capitalize on high investor interest. This move marks the company's first major bond issuance since 2021 as it seeks to fund its aggressive AI infrastructure expansion.
Ars Technica · 96d ago
AMD has acquired MEXT, a company specializing in AI memory optimization, to reduce data center TCO. The move aims to address the surging memory demands of Agentic AI and next-gen applications.
cnBeta (Full RSS) · 96d ago
Kazakhstan has signed a $10 billion agreement with Firebird to construct a 'Data Center Valley'. This initiative aims to transform the oil-rich nation into a regional computing hub for artificial intelligence.
The Next Web (TNW) · 96d ago
Schneider Electric and Foxconn have formed a strategic partnership to design and scale next-generation AI data centers. The collaboration leverages Schneider's energy management expertise and Foxconn's manufacturing scale.
The Next Web (TNW) · 96d ago
SpaceX has rented its Colossus 1 data center to Anthropic after failing to resolve latency issues for its own Grok AI models. The Memphis-based facility could not be effectively integrated with SpaceX's other data center campuses.
The Next Web (TNW) · 99d ago
A transmission fault in Ashburn, Virginia, reportedly removed more than 3 gigawatts of data-center load from the grid in seconds. The incident follows an earlier failure that affected about 60 facilities, highlighting the fragility of concentrated AI infrastructure.
MIT Technology Review · 9d ago
Arm has launched Neoverse CSS N4, a server CPU platform designed to address rising compute demand in the agentic AI era. Each die can support up to 128 cores.
InfoQ中国 · 11d ago
Broadcom reported fiscal Q3 2026 revenue of $29.591 billion, up 86% year over year, while net income rose 216% to $13.088 billion. AI semiconductor revenue reached $16.7 billion, and the company expects $21.7 billion in AI semiconductor revenue in Q4.
IT之家 · 17d ago
Hewlett Packard Enterprise raised its sales outlook for the current and following fiscal years. The company attributed the stronger outlook to demand for servers and networking equipment supporting AI workloads.
Bloomberg Technology · 17d ago
AI demand is turning compute rental into a seller’s market, with high-end GPUs, data centers, power, networking, and operations all in short supply. Profitability varies widely by customer segment and asset model, while capital providers and operators with long-term contracts capture much of the value.
虎嗅 · 19d ago
The Microsoft-linked DataOne data center project in Vineland, New Jersey, is facing community backlash over alleged unpermitted construction and environmental concerns. Complaints involve 62 gas turbines, a 1.5-million-gallon LNG tank, noise pollution, and potential water issues at the $19.4 billion facility.
Tom's Hardware · 21d ago
A Microsoft-backed AI data center has allegedly operated dozens of large generators without the required permit. The accusation raises regulatory and environmental compliance concerns for AI infrastructure projects.
Engadget · 22d ago
SpaceXAI plans to use NVIDIA Vera CPUs for Grok infrastructure and adapt a Vera Rubin NVL72 system for its planned Starmind AI satellite. The proposal signals an expansion of NVIDIA-based computing across terrestrial and space-oriented AI infrastructure.
TestingCatalog · 25d ago
Nvidia’s upcoming earnings report will test whether its AI momentum can meet exceptionally high investor expectations. Beyond its dominant data center business, the company must convince investors it can remain the leading AI platform for years to come.
Bloomberg Technology · 25d ago
AI server racks now generate around 120 kW of heat, making traditional air cooling insufficient for high-density computing. The article examines cold-plate liquid cooling, quick connectors, coolant distribution units, rear-door heat exchangers, and two-phase immersion cooling as the industry’s main responses.
虎嗅 · 26d ago
文章称英伟达 AI 服务器可能因内存成本飙升而涨价约 15%。对于 1GW 数据中心，这一硬件成本变化可能带来高达 50 亿美元的额外支出。
量子位 · 27d ago
Nvidia has reportedly notified major customers that servers containing its AI chips will become more than 15% more expensive in many cases. The increases are expected to apply to systems shipped early next year, including configurations using Vera Rubin and Grace Blackwell chips, amid soaring memory costs.
SCMP Technology · 28d ago
Cerebras Systems introduced a new, faster computer powered by its own AI chips. The company says the system widens its performance advantage over Nvidia-based equipment.
Bloomberg Technology · 31d ago
NVIDIA and OpenAI are reportedly nearing an agreement to finance a large data center campus in Ohio. The revised structure would have NVIDIA guarantee only half of the planned financing, rather than the full amount initially considered.
cnBeta (Full RSS) · 35d ago
Reports from Bank of America and UBS indicate that NVIDIA's upcoming Vera Rubin Ultra Kyber rack could contain 340.4TB of total memory. Rising memory prices and architectural changes reportedly push LPDDR5X procurement costs above those of HBM4E, with the rack's total build cost potentially reaching $41.6 million.
cnBeta (Full RSS) · 38d ago
China’s server vendors are being reordered as more capital flows into AI infrastructure. However, not all AI infrastructure spending reaches traditional server manufacturers.
钛媒体 · 39d ago
Point2 Technology has raised $136 million in total Series B financing to commercialise RF cables designed for AI data centres. The company says the cables offer greater reach than copper and lower latency than optical connections.
The Next Web (TNW) · 40d ago
Intel’s flagship Diamond Rapids server processor, also known as Xeon 7, is reportedly being developed with up to 256 performance cores. The chip is intended to compete directly with AMD’s next-generation EPYC Venice platform.
cnBeta (Full RSS) · 41d ago
Samsung Heavy Industries and Dallas-based Mousterian signed an engineering agreement to design a floating data center, with the first project planned for Texas. The concept aims to address AI infrastructure constraints around land and water.
The Next Web (TNW) · 44d ago
耐士勞新任總裁李建提出三項戰略：整合戶外庭院品牌、自研激光雷達，以及布局四足機器人。公司希望從產品導向轉型為由技術與場景驅動的機器人生态企業。
钛媒体 · 16d ago
Sunshine Digital Technology plans to invest up to 980 million yuan to build the Sunshine Intelligent Computing Center in Xingning, Guangdong. The facility is expected to begin operations in October 2027, but the company highlights approval, financing, cross-industry, and continued-loss risks.
36氪 · 47d ago