memory

Google Research unveils TurboQuant, a training-free compression algorithm that reduces LLM key-value cache memory by 6x and boosts inference throughput up to 8x, sending SK Hynix and Samsung shares down 5-6 percent.

Micron Technology posted record fiscal Q2 revenue of 23.9 billion dollars, nearly tripling year over year on surging AI memory demand, while guiding fiscal Q3 to 33.5 billion dollars -- a single quarter that would exceed its entire annual revenue through fiscal 2024.

Samsung begins commercial HBM4 shipments with 3 TB/s bandwidth per stack, gaining a timing edge over SK Hynix as memory becomes the critical bottleneck in AI infrastructure.