A new Chinese law increases scrutiny of Taiwanese investments, according to DIGITIMES. The move may affect the semiconductor supply chain, potentially impacting availability and costs of hardware used for on-premise LLM inference and training.
Taiwan scooter sales rose in June, with SYM holding its lead and Gogoro surging. A sign of industrial vitality that also impacts on-premise AI hardware, where component availability and local supply chain resilience are increasingly strategic.
Rumors point to Meta exploring an AI cloud offering to monetize its massive infrastructure investments. It is not an AI retreat but a strategy to generate returns. The move reignites the debate about when to build an on-premise stack versus relying on external services, balancing TCO, control, and data sovereignty.
The Chinese giant has reportedly blocked Anthropic's coding assistant over data protection fears. The move underscores how technological sovereignty is reshaping corporate policies, with direct implications for those evaluating on-premise AI stacks today.
The development team has published the second release candidate for the Linux 7.2 kernel, with Linus Torvalds noting 'Things look very normal.' A signal of a mature development cycle, highly relevant for on-premise infrastructures that depend on stability. The stable release is expected in August.
Micron has started construction on an expansion of its Hiroshima plant, aiming to increase production of memory chips for artificial intelligence. The move addresses the growing demand for high-bandwidth memory (HBM) essential for GPUs and accelerators used in LLM training and inference, signaling the hardware industry's readiness for wider AI workloads, including on-premise deployments.
CG Semi's packaging and test plant in Gujarat has begun commercial production, marking India's third semiconductor hub. Beyond electronics, it’s a building block for on-premise AI and digital sovereignty, where local chip availability cuts risk, lead times, and costs.
