AI-driven memory demand is growing exponentially, pushing the industry into a structural supply shortage. As supply tightens and contract prices surge in 2026, major AI companies and cloud service providers (CSPs) are increasingly signing long-term agreements (LTAs) lasting three to five years with memory manufacturers to secure supply and manage soaring costs.

China's leading DRAM manufacturer, ChangXin Memory Technologies (CXMT), expects to report net profit of CNY131 billion to CNY132.5 billion (≈ US$19.58 billion to US$19.8 billion) for the first nine months of 2026, reversing a year-earlier loss as surging AI computing demand fuels memory chip shortages and higher prices.

South Korean semiconductor equipment maker Hanmi Semiconductor showcased its high-bandwidth memory (HBM) and AI chip packaging equipment at SEDEX 2026, including the TC Bonder 4 for HBM4 production and the Wide TC Bonder, scheduled for launch in 2027. The company also introduced five 2.5D packaging systems designed to accommodate increasingly large AI chip packages.

South Korean ASIC design solutions provider ASICLAND showcased its proprietary AxHub AI chip design and verification platform at SEDEX 2026 as it seeks to accelerate custom chip development and strengthen its advanced process and packaging capabilities through its Taiwan R&D center. Memory controller projects remain its primary revenue source, supporting its expansion into AI chip mass production and overseas markets.

Malaysia's push to move from assembling chips to owning the designs behind them has reached the point where it must find buyers. On October 16, the government unveiled six semiconductor innovations developed by local companies, putting their combined sales potential at MYR300 million (approximately US$73.5 million), while Economy Minister Akmal Nasrullah Mohd Nasir said winning customers and passing industry qualification were still ahead.

Micron is developing a memory architecture for humanoid robots, joining Samsung Electronics and SK Hynix in efforts to serve AI systems that must process sensor data while operating within battery power limits. The companies are pursuing both memory designs and storage products for robots and autonomous vehicles.

Infineon Technologies has lined up 8-inch silicon carbide (SiC) wafers from Wolfspeed under an expanded and extended long-term supply agreement. The deal gives the German chipmaker another source of the larger substrates it needs to convert its SiC fabs. It gives the US materials supplier a commitment from a top-tier customer about a year after its financial restructuring.

Academia Sinica academician Chenming Hu has published an autobiography titled Find the Solutions, taken from the words on a crystal plaque presented to him by then TSMC chairman Morris Chang. Hu, who has worked in both academia and industry, said universities' ability to conduct thorough testing could help the semiconductor industry overcome technological bottlenecks.

AI data center investment continues to expand, keeping the global memory market undersupplied in the fourth quarter of 2026. Rising DRAM and NAND Flash contract prices are boosting expectations for Samsung Electronics and SK Hynix in 2027, although weak consumer demand and long-term agreement (LTA) price caps will limit further increases.

This week's coverage (October 12 to October 18, 2026) highlights semiconductor capacity, AI infrastructure, and the commercialization challenges behind ambitious technology strategies. Tata's Dholera fab has early customer commitments but must convert them into qualified, recurring wafer production while facing shareholder scrutiny. Intel is rebuilding through process execution, partnerships, and advanced packaging. Micron and TSMC are expanding AI-related capacity in Taiwan and Japan, while companies including Murata, AWS, Hitachi, and XPeng advance physical AI, robotics, and engineering exports.

Nvidia's custom memory design is shifting control of how high-bandwidth memory (HBM) is built from memory makers to its largest customer, and steering a key part of that work to TSMC. Samsung Electronics, which has made every part of its HBM stacks in-house, has agreed after an internal review to use TSMC's 3nm process for the base die of NVHBM, Nvidia's next-generation custom HBM, the Seoul Economic Daily reported on October 18, citing industry sources. The newspaper said it would be the first time the two companies have worked together at the HBM manufacturing stage.