The storage market is experiencing a significant boom driven by the explosive demand for AI applications, particularly in high-bandwidth memory (HBM) used in AI GPUs. The demand for AI inference has surged, now accounting for 66% of AI compute, reversing the previous dominance of training. This shift has led to a critical bottleneck in HBM bandwidth and capacity, essential for efficient AI inference. Major players like Samsung, SK Hynix, and Micron are at the forefront of HBM production, but the complex manufacturing process and high demand have led to severe supply constraints. HBM production requires significant wafer capacity, impacting traditional DRAM and NAND supplies. Despite efforts to expand capacity, the 2-3 year cycle for building new semiconductor fabs means supply will remain tight until at least 2027. The current supercycle in the memory chip market is marked by unprecedented demand and low inventory levels, with Goldman Sachs forecasting a 4.9% DRAM supply deficit in 2026. This imbalance is driving significant price increases and long-term supply agreements, as manufacturers prioritize high-margin HBM production over traditional memory products.