TLDR
- Google introduced TurboQuant, a compression algorithm capable of cutting AI memory needs by up to six-fold
- Samsung experienced a 4.8% decline while SK Hynix saw a 6.23% drop following the announcement
- American memory chip manufacturers including Micron, SanDisk, and Western Digital declined between 1.6% and 3.5%
- Market experts characterize the downturn as temporary profit-taking rather than indication of weakening long-term demand
- TurboQuant will be formally presented by Google at the ICLR 2026 conference scheduled for April
Shares of South Korean memory chip manufacturers Samsung and SK Hynix experienced substantial declines on Thursday following Google researchers’ announcement of TurboQuant, a breakthrough AI memory compression algorithm.
Samsung’s stock price decreased by 4.8%, while SK Hynix witnessed a steeper 6.23% fall during trading on the Korea Exchange. The two companies ranked among the leading contributors to the KOSPI index’s decline, which reached as high as 3% during the trading session.
The broader electrical and electronics sector index experienced a 4.76% decline, settling at 3,592.22 by market close.
The downturn in Asian markets mirrored losses experienced by American memory chip companies the previous day. Micron Technology shares fell 3.40%, SanDisk decreased 3.50%, and Western Digital slid 1.63% during New York Stock Exchange trading.
Google’s research team introduced TurboQuant earlier in the week. The innovative algorithm enables artificial intelligence models to process up to six times greater capacity while utilizing identical memory resources.
According to Google, the technology offers optimization capabilities for vector search functions that drive prominent search platforms.
Market participants expressed concerns that reduced memory requirements might diminish demand for cutting-edge memory semiconductors that Samsung and SK Hynix provide to data center operators.
Both manufacturers have experienced significant benefits from AI-related demand growth. This demand trajectory had been projected to create memory chip supply constraints in recent quarters and powered substantial stock price appreciation for both companies.
Market Analysts Challenge Pessimistic Interpretation
The majority of financial institutions interpret the decline as temporary profit realization. Analysts contend the technology may actually bolster memory demand in the longer term.
Surim Lee, an equity research analyst at DS Investment & Securities, explained that technologies designed to reduce memory consumption typically expand aggregate demand rather than diminish it. He emphasized that TurboQuant’s development stems from memory constraints in AI infrastructure reaching critical thresholds.
Lee further noted that efficiency improvements generate a cyclical pattern where cost reductions drive increased utilization and capital reinvestment, rather than decreased overall demand.
Han Ji-young from Kiwoom Securities observed that as AI models achieve greater efficiency and capability, paradoxically, total AI demand could experience growth.
Han also suggested the pronounced stock decline likely reflected investor weariness following the early-year memory price surge, with the TurboQuant announcement providing justification for profit-taking.
Historical Trends Reinforce Long-Term Perspective
Historical evidence demonstrates that innovative technologies enhancing resource efficiency have frequently catalyzed accelerated industry expansion and elevated aggregate demand, contrary to reduction.
Heo Jae-hwan from Eugene Investment & Securities stated that server and semiconductor segments connected to artificial intelligence, where supply constraints already exist, will likely demonstrate greater resilience compared to alternative industries.
Google announced plans to formally unveil TurboQuant at the ICLR 2026 conference scheduled for April.
Samsung and SK Hynix concluded trading at 1,801,000 won and 9,330,000 won respectively on the Korea Exchange on March 26.





