Google's effort to expand its tensor processing units (TPU) beyond its own cloud is meeting resistance from some of the AI infrastructure companies best positioned to distribute alternative chips, with executives from Nebius, Lambda, and CoreWeave saying they do not plan to adopt TPUs anytime soon, according to The Information.
As power constraints and tighter policy controls in Singapore limit expansion, Malaysia is rapidly emerging as Southeast Asia's most important data center hub. A DIGITIMES research report finds that Chinese cloud service providers (CSPs), including ByteDance and Alibaba, are accelerating their shift southward, using Malaysia's unique supply chain pathways to deploy high-end AI computing power locally, including Nvidia's B200 systems.
The transition from 800G to 1.6T optical modules is no longer an upgrade cycle — it is a physics-driven inflection point.
OpenAI is reworking its US$500 billion Stargate initiative, shifting from a fixed data center joint venture toward a more flexible strategy for securing computing capacity as demand for artificial intelligence infrastructure accelerates.
Lite-On Technology reported first-quarter revenue of NT$43.4 billion (approx. US$1.35 billion), up 19% from a year earlier, as surging demand for AI infrastructure fueled rapid growth in its cloud-related business.
OpenAI's aggressive push to secure AI computing capacity is facing growing scrutiny, with signs that its multibillion-dollar data center procurement plans may be reassessed, reportedly raising questions for a supply chain led by Quanta Computer, Wistron and Foxconn.
Prices for high-end AI servers in China are diverging sharply from global benchmarks, with systems built around Nvidia's B300 chips now fetching scarcity-driven premiums that reflect tightening export controls and surging domestic demand.
Microsoft executives defended the company's rising capital spending after reporting stronger-than-expected fiscal third-quarter results, saying demand for Azure, Copilot, and AI infrastructure continues to exceed available capacity.
Alphabet's first-quarter 2026 results point to a company increasingly defined by AI-driven momentum across search, cloud, and subscriptions—with CEO Sundar Pichai framing the quarter as evidence that its "full-stack" AI strategy is beginning to translate into measurable business performance.
China's AI infrastructure race is entering a phase where scale alone is no longer enough. As large models move from training to real-world deployment, the focus is shifting toward cost efficiency, system integration and application delivery. Against this backdrop, Unisplendour is emerging as a key beneficiary of the buildout, though its long-term trajectory will depend on how effectively it converts infrastructure strength into sustainable commercial value.
