SpaceX's multibillion-dollar cloud agreement with Google underscores a growing shift in the AI industry from building proprietary models to monetizing computing infrastructure. The deal not only secures a major recurring revenue stream ahead of SpaceX's IPO but also highlights persistent demand for AI capacity as technology companies race to meet surging enterprise adoption.

Below are the most-read DIGITIMES Asia stories from the week of June 1-7, 2026:

Nvidia and SK Hynix formally announced a multiyear technology partnership on June 7 at SK's Seorin Building in Seoul, during Jensen Huang's third public meeting with SK Group chairman Chey Tae-won on this Korea trip. The agreement covers next-generation memory co-development across Nvidia's full product roadmap and extends into semiconductor design, AI factory infrastructure and digital manufacturing.

Nvidia and Doosan Group are widening their collaboration to develop physical AI, robotics, and AI factory infrastructure that could shape industrial automation worldwide. The effort spans robotics, heavy equipment, power systems, and advanced materials, highlighting how global AI growth is increasingly tied to manufacturing, energy, and data center supply chains.

According to Nvidia's press release, SK Telecom plans to build a gigawatt-scale AI cloud in Korea, with the first AI factory set to go online in 2027. The project signals how telecom operators may evolve into global AI infrastructure providers, shaping access to computing capacity, energy use, and industrial AI deployment.

KT Cloud is accelerating its artificial intelligence data center rollout in South Korea with modular construction, alternative power sourcing, and a digital twin operations platform. The move highlights how long permitting timelines, grid limits, and evolving GPU demands could shape AI infrastructure availability for users far beyond Korea.

Dassault Systèmes, Quanta Cloud Technology, and Nvidia are advancing industrial digital twins for AI factories, a shift that could shape how companies worldwide build, run, and scale generative AI systems. The effort targets more efficient inference, stronger lifecycle control, and lower energy use as demand for accelerated computing rises globally.

France secured more than EUR110 billion (US$126.8 billion) in artificial intelligence and data center investment commitments at the recent Choose France summit, with the largest single pledge totaling EUR75 billion for three northern data centers, but officials and investors warned that grid constraints and administrative approvals could delay delivery. According to Financial Times and Reuters, the summit produced a record level of investment promises that together aim to add about 10 GW of computing capacity.

As hyperscale cloud providers and enterprise customers race to build increasingly powerful artificial intelligence infrastructure, power delivery and energy efficiency are emerging as critical constraints. Industry analysts expect AI rack power requirements to exceed 1 MW per rack in the coming years.

South Korea's private data center market is set to expand rapidly as generative AI demand drives a global investment boom, with AI infrastructure, power supply, geographic dispersion, and data sovereignty reshaping the industry's development path. Jong-hyun Shin, a researcher at the Korea Data Center Council (KDCC), said 2028 is expected to be the peak year for data centers finishing construction.

Aspeed chairman Chris Lin said on June 3 that rapid growth in agentic AI is boosting demand for AI and general-purpose servers, with the chip maker's main constraint now supply-chain capacity rather than orders. He said the company already has visibility into a large volume of 2027 orders as customers rush to secure supply.

Taiwan Mobile announced at Computex 2026 that it signed a memorandum of understanding (MOU) with GMI Cloud to expand AI data center and high-end compute services across emerging markets in Southeast Asia, Northeast Asia, the Americas, and Europe. The firms said the agreement aims to deliver AI data center offerings that comply with local regulations and cybersecurity standards while competing for cross-border compute business.

AI rack power density has increased 50 times in five years — and the next jump may be the hardest yet.

As artificial intelligence fuels demand across the electronics supply chain, Taiwanese conductive paste supplier Ample Electronic is seeing a rebound in one of the industry's most overlooked segments: passive components.

Intel and Foxconn have signed a memorandum of understanding (MoU) to cooperate on AI rack infrastructure, edge AI, physical AI platforms, and custom chip design services. The deal highlights Intel's bid to regain ground in artificial intelligence (AI), while broadening Foxconn's role in the global AI supply chain.

Compal Electronics is accelerating a shift away from its long-standing PC business as AI servers and other non-PC lines gain momentum. The move could have broad implications for global supply chains, data center capacity, and technology customers seeking more diversified hardware partners across Asia, the US, and Vietnam.

Foxconn and Intel have announced a strategic partnership focused on AI racks, Edge AI, and Physical AI. The move signals Intel's effort to rebuild competitiveness in a market increasingly shaped by Nvidia's dominance in AI training and inference.

As AI pushes liquid cooling into mainstream server thermal design, suppliers are widening their focus beyond Nvidia's benchmark thermal designs. Industry players say pumps are emerging as a key lever for improving cooling efficiency, alongside flow rate and fluid speed.

As the global demand for AI accelerates, a new industrial challenge has emerged: the sheer scale of energy required to power the "AI factories" of the future. This concern was the centerpiece of presentations by several industry figures at Computex in Taipei, where they presented solutions under development, from raw material innovations to shrinking component sizes.

Foxconn announced on June 4 a strategic partnership with Intel to explore end-to-end AI solutions spanning chips, racks, systems, and applications, with cooperation covering AI rack infrastructure, edge AI, and physical AI platforms. The partnership will also extend to design services for custom ASICs, SoCs, and system integration, combining Intel's processor, silicon photonics, and software ecosystem with Foxconn's global manufacturing scale, system integration capabilities, and AI data center deployment experience.

Foxconn and Foxconn Industrial Internet (FII) said at Computex that production of Nvidia's next-generation AI products is progressing smoothly, with shipments expected before the end of the third quarter of 2026. The outlook matters globally because it signals faster AI hardware deployment, lower computing costs, and wider adoption across consumer, enterprise, and automotive markets.

Server chassis maker Chenbro is continuing to evolve, aiming to perfect the craftsmanship of mechanical components to become the world's top rack supplier, according to company CEO Corona Chen.

Global AI computing demand is sweeping through the infrastructure market. Kentec, the AI data center (AIDC) system integrator and turnkey solution provider under Kenmec Mechanical Engineering, expects first-half 2026 revenue to rise more than 200% year on year, with the real surge concentrated in the second half of the year.