Crypto World News reports that Citrini Research states AMD and Apple are advancing flash memory replacement DRAM solutions in AI products to reduce memory costs. AMD is acquiring MEXT to optimize flash memory performance, while Apple is achieving device-side optimization through LLM in a Flash technology. The demand for KV caching in AI inference is increasing, and HBM occupies 25% of DRAM capacity, while flash memory costs only 1/55 of DRAM. Through controller optimization, NAND stacking, and cell mode adjustments, a capacity and bandwidth alternative can be provided for edge AI.