Micron perps slide about 6.5% in eight minutes after DeepSeek Flash model news

A memory-efficiency claim from a Chinese AI lab rattled Micron traders in a market that never closes Micron Technology perpetual futures dropped approximately 6.5% in eight minutes over the October 17-18 weekend. The trigger was DeepSeek and its Flash model, which is pitched on doing more with less memory. The pressure traces back to DeepSeek’s V4.1 Flash model, released on September 10, 2026. DeepSeek claimed it cuts KV-cache high-bandwidth memory (HBM) usage by 75% compared with the previous version. A quick translation helps. The KV-cache is the model’s short-term memory. It is a running notepad of everything said so far, so the model does not reread the whole conversation before each word. That notepad usually sits in HBM, the expensive, ultra-fast memory stacked next to AI chips.

BeCrypto News shows a short excerpt and links to the publisher. The full report belongs to Crypto Briefing.