Micron Technology Slides 23% as Google’s AI Cuts Memory Needs Sixfold
Google’s TurboQuant technique cuts memory requirements for large language model inference by at least 6x, triggering a two-day selloff in flash-focused stocks while DRAM suppliers like Micron Technology remained stable. Micron then extended a six-session decline, shedding 23% as investors question the sustainability of AI memory demand and pricing.
1. AI Memory Efficiency Breakthrough
Google’s TurboQuant technique reduces memory use and data movement for large language model inference by at least a factor of six, reshaping demand expectations across the memory ecosystem.
2. Segmented Market Reaction
Flash and storage-focused issuers like Kioxia plunged over two days, while high-bandwidth memory and DRAM suppliers such as Micron Technology and SK Hynix held gains after investors distinguished between GPU-centric memory and general storage demand.
3. Micron’s Extended Slide
Micron Technology has tumbled approximately 23% over six sessions as traders weigh the durability of AI-driven memory price increases, competitive pressures and supply dynamics on the company’s data center revenue.
4. Outlook on Memory Demand
Analysts at major firms note core DRAM tied to compute-intensive workloads may remain resilient, but waning flash requirements could pressure NAND revenues, with memory price trends and hyperscaler orders under close watch for future guidance.