Taiwanese memory chipmaker Nanya Technology has warned that the global memory shortage is likely to continue through at least 2027 as artificial intelligence-driven demand continues to outpace supply growth across the semiconductor industry. The company’s outlook reflects growing concerns among major chip manufacturers about prolonged constraints in the global DRAM and high-bandwidth memory (HBM) markets.
According to industry reports, the rapid expansion of AI infrastructure, cloud computing, and data center investments has dramatically increased demand for advanced memory chips. AI systems require enormous volumes of high-speed DRAM and HBM to support large language models, generative AI workloads, and high-performance computing applications. However, manufacturing capacity expansion has not kept pace with the surge in demand.
Nanya reportedly expects memory supply shortages to persist through the first half of 2027, with demand continuing to exceed available production capacity. Industry estimates suggest that current memory production growth rates remain significantly below what would be required to fully meet global AI-driven demand. Reports indicate that suppliers may only be able to satisfy around 60% of projected AI memory demand by 2027 if current expansion trends continue.
The shortage is also being intensified by the semiconductor industry’s strategic shift toward high-bandwidth memory production. Major players including Samsung, SK hynix, and Micron have increasingly prioritized HBM manufacturing because of its higher profitability and strong demand from AI accelerator companies such as NVIDIA. This shift has reduced capacity available for conventional DRAM products used in PCs, smartphones, and consumer electronics.
Industry leaders across the semiconductor ecosystem have echoed similar concerns in recent months. Samsung executives recently warned that memory shortages could worsen in 2027 as AI demand continues to surge globally. Analysts believe the supply-demand imbalance is becoming more structural rather than cyclical, driven by long-term AI infrastructure expansion and limited fabrication capacity growth.
The prolonged shortage has already begun impacting broader technology markets. Rising memory prices have increased costs for smartphones, PCs, servers, and enterprise hardware, while some manufacturers have reported difficulties securing long-term supply contracts. Market research firms have also projected declines in global smartphone and PC shipments due to increasing component costs and constrained supply availability.
To address future demand, memory manufacturers are investing heavily in new fabrication facilities and advanced semiconductor packaging technologies. However, analysts note that new semiconductor plants require several years to become fully operational, meaning meaningful supply relief may not arrive until late 2027 or even 2028. Until then, AI-driven demand is expected to continue reshaping the global semiconductor and memory markets.
