AMD said it will acquire MEXT, a move aimed at strengthening its AI and data center portfolio amid rising global memory demand. The deal is intended to help customers improve performance, reduce infrastructure costs, and accelerate the deployment of AI, analytics, and high-performance workloads across cloud and enterprise systems.
China has recently eased controls on some indium phosphide (InP) substrates, relieving a bottleneck in optical communications capacity for the second half of the year. But supply chain players say the long-term priority is still to expand substrate capacity from non-China sources, with supply security for the AI industry outweighing price.
Salesforce has agreed to acquire Fin, a customer agent company formerly known as Intercom, in a deal that could reshape how businesses worldwide automate support. The purchase highlights growing competition in AI-driven service tools, with implications for companies seeking faster responses, lower costs, and broader deployment across global customer channels.
Nvidia's dominance in AI is extending beyond model training and deeper into inference—the fast-growing segment of the AI market responsible for running deployed models and generating revenue.
The deal between Gautam Adani's group and American engineering firm Jabil signals a broader shift toward domestic AI infrastructure in India, with potential consequences for global technology supply chains, foreign cloud dependency, and sovereign technology ambitions. The partnership aims to accelerate capacity, reduce reliance on external providers, and attract investment.
Indian regulators allege wastewater contamination at a key Apple supplier site in Tamil Nadu, raising fresh questions over environmental compliance and the stability of India's fast-expanding iPhone manufacturing ecosystem.
Foxconn and Schneider Electric announced a strategic partnership to jointly develop a next-generation data center reference architecture, combining Foxconn's capabilities in advanced computing platforms, server rack integration, and global manufacturing with Schneider Electric's expertise in power systems, cooling technology, and energy management. The collaboration aims to help customers build and operate large-scale computing infrastructure faster, more efficiently, and with greater predictability. Production for the partnership is expected to begin later in 2026.
Yaskawa Electric is betting heavily on physical AI, the field that enables robots to operate autonomously, as it positions itself for a boom in AI robots. Since starting work with Nvidia in 2023, the company has completed proof-of-concept tests with more than 100 companies and is now seeking an effective commercial model.
After Apple's WWDC 2026 keynote, online rumors claimed that only premium devices with 12GB of memory could run on-device AI, but DIGITIMES analyst Luke Lin said on a podcast that that was just Apple's messaging. Most on-device AI features only need 8GB of memory; only the Apple-defined advanced on-device AI requires the higher 12GB spec.
Physical AI is emerging as a new frontier of model development. Any model, however, is only as good as the data used to train it. Because of this, Japanese startup APTO is launching a physical AI infrastructure lab to help plug the data gap needed to create vision-language-action (VLA) models, with a focus on imitation learning.
SAP reported that enterprise AI agents are shifting from proof of concept to real-world deployment as companies aim to embed generative AI into operational workflows such as finance and supply chain management rather than limiting it to chat and summarization tools. The firm said many AI applications reached about 80% accuracy during the proof of concept stage, but that core workflows require substantially higher reliability and human oversight to ensure safety and compliance.
More coverage
