Memory Chipmakers Poised for Growth as AI Demand Surges, Predicts Citi
Leading financial institution Citigroup (Citi) foresees a bright future for memory chipmakers, driven by the rising demand for artificial intelligence (AI) applications. Citi predicts that specialized memory and storage products intended for AI chips will gain a larger market share, reducing volatility in the sector. In the first quarter of 2023, inventory levels for memory chips, such as generic dynamic random-access memory (DRAMs), reached multi-year highs. Elevated stock levels typically lead to reduced pricing power for manufacturers. However, Citi estimates that inventories have bottomed out after declining by 33% since early 2022. As the market recovers and diversifies to produce more advanced and complex chips for AI applications, the investment bank expects earnings at chipmakers to rise.
Citi analysts led by Peter Lee explained in a note to clients on January 1 that they anticipate memory products to become semi-customized as AI memory products lead to increased product diversification and complexity. They pointed out that the memory market is likely to veer towards higher product optimization to meet customer needs, similar to the foundry market. In light of this trend, Citi recommended investments in two specialized memory chips: high bandwidth memory (HBM) and multi-ranked buffered DIMMs (MRDIMM). These chips are expected to be widely adopted due to their suitability for AI applications. HBM provides greater bandwidth to accelerate parallel computing, while MRDIMM supports expanded memory capacity. Citi predicts that these chips will grow from comprising 3.8% of total DRAM shipments in 2024 to 8.8% in 2027. Therefore, chipmakers SK Hynix and Micron are likely to benefit from the growth of these semi-customized memory chips. Citi projects that SK Hynix shares could rise by 63% and Micron Technology shares by 11% over the next 12 months.
Moreover, the expansion of AI processing beyond centralized server farms and into consumer devices like PCs and mobile phones presents additional opportunities. Citi forecasts that new memory chips will emerge to enable on-device AI, providing low power consumption and high bandwidth. Dell is notably one of the most active companies in developing next-generation memory chips for thin laptops. Referred to as Compression Attached Memory Module, these memory chips are expected to contribute to a 9.8% increase in Dell’s shares to $84 over the next 12 months.
With the increasing demand for AI across various industries, memory chipmakers are poised for growth. Citi’s predictions align with the evolving market landscape and highlight the potential for specialized memory chips to become dominant players. As both AI technology and memory chip offerings continue to advance, investors may find these insights valuable in their decision-making process.
