Samsung Electronics has reportedly secured new AI chip manufacturing orders from major US cloud service providers for a second consecutive quarter, with commercial shipments expected to begin in 2028 as tight capacity at TSMC creates opportunities for Samsung's foundry business.

Chinese DRAM maker ChangXin Memory Technologies (CXMT) is expanding into higher-value memory markets, introducing DDR5 server modules, mass-produced LPDDR6 chips, and automotive DRAM at IC China 2026, which opened on November 12. The expansion, backed by manufacturing advances and higher research spending, signals CXMT's ambitions beyond its established mobile memory business.

SK Hynix is working with Intel and other industry partners to develop memory-centric computing architectures that would allow AI servers to share memory resources, as agentic AI drives up demand for memory capacity and bandwidth.

Intel is courting South Korea's AI ecosystem on two fronts: a 23-company alliance built around Korean-made servers and an inference GPU that swaps costly high-bandwidth memory (HBM) for low-power DRAM. At the 2026 Intel AI Summit Seoul, held at COEX on November 12, the US chipmaker also offered its custom chip design and foundry services to a market crowded with neural processing unit (NPU) startups and home to two of the world's largest memory makers.

Megaspeed International said it is reviewing its graphics processing unit inventory and export-control practices as it seeks to demonstrate compliance with applicable US laws. The company's statement, issued November 12, 2026, signals scrutiny of technology shipments and regulatory obligations, with relevance for businesses and governments that operate across markets.

The battle for control of Dutch chipmaker Nexperia is intensifying after a Dutch court further restricted Chinese parent Wingtech Technology and related parties from interfering with the company's operations. Meanwhile, Nexperia China is accelerating efforts to establish an independent supply chain, deepening the divide between the two sides' management and manufacturing operations.

PsiQuantum and AMD are expanding their partnership to develop infrastructure for utility-scale, fault-tolerant quantum computers and hybrid systems combining quantum, high-performance, and AI computing. The agreement includes AMD processors, GPUs, software, and an equity investment, with potential applications in drug discovery and other industries worldwide as systems scale globally.

Special Melted Products, a UK subsidiary of Taiwan-based wire, cable, and stainless steel manufacturer Walsin Lihwa, is partnering with National Taiwan University (NTU) and the University of Sheffield to establish an advanced international research center. The initiative will combine Sheffield's traditional steelmaking base with strengths in heterogeneous integration and compound semiconductors, targeting the aerospace and defense industries as key commercial outlets.

Is Nvidia's moat still intact? A slide in Anthropic's IPO filing has quickly sparked industry debate after revealing that Anthropic's Claude Mythos Preview model spent only about three days exploring how to run inference for another AI model on Advanced Micro Devices (AMD) Instinct MI355X GPUs.

For generations, Sheffield's industrial identity has been forged in steel. Yet behind that familiar legacy, a less visible technological transformation has been unfolding for decades.

Imagine a maintenance robot operating inside a fusion reactor, exposed to temperatures reaching 600 degrees Celsius and intense radiation, with an expected service life of 25 years. Developing electronics capable of surviving such conditions would be an extraordinary engineering achievement. But manufacturing the chips is only half the problem.