Key Takeaways
- Memory chip stocks including Micron and SanDisk tumbled following Google’s announcement of TurboQuant, an algorithm that compresses AI memory requirements by 6x or more
- Market participants worried the innovation would slash memory chip demand, driving both equities down more than 15% from their recent peaks
- Mizuho’s Vijay Rakesh contends the market reaction is excessive and maintains Outperform recommendations for both companies
- Rakesh contends that efficiency breakthroughs like TurboQuant typically drive greater AI implementation, ultimately increasing memory demand through Jevons’ Paradox
- AI server NAND content has grown twofold over the last twelve months, while spot market prices continue their quarterly ascent
Shares of Micron and SanDisk experienced significant declines last week following Google’s publication of research detailing TurboQuant, a novel compression technique that reduces AI model memory requirements by a minimum factor of six. The technology simultaneously delivers inference speed improvements of up to eight times while maintaining model accuracy.
Investors interpreted the development as bearish for memory semiconductor manufacturers. The logic was straightforward: if AI models require less memory per deployment, demand for chips from producers like Micron and SanDisk would decline proportionally.
Both equities have retreated at least 15% from all-time highs established in late last month. Thursday’s trading session saw SanDisk decline 5.9% to $652, while Micron fell 5.5% to reach $347.78.
Compounding market anxieties, President Trump’s Wednesday evening remarks failed to provide clarity regarding the timeline for resolving the Iran situation, contributing to cautious sentiment that carried into Thursday’s trading.
Google researchers initially examined TurboQuant concepts in 2025, with updated findings on AI inference capabilities released in subsequent publications.
Mizuho’s Case for Buying the Weakness
Vijay Rakesh, an analyst at Mizuho Securities, challenged the market’s interpretation in research distributed to institutional clients. He reaffirmed Outperform ratings for both Micron and SanDisk while maintaining price objectives of $530 and $710 respectively.
Rakesh advised clients to “buy the TurboQuant memory pullback,” characterizing concerns about peak memory consumption as “overblown.”
His central thesis rests on historical patterns showing that AI efficiency gains consistently generate increased expenditure rather than reduced spending. This phenomenon, termed Jevons’ Paradox in economic theory, describes how improvements in resource efficiency paradoxically drive higher aggregate consumption of that resource.
Rakesh cited three historical precedents. Server virtualization technology was anticipated to reduce physical server requirements but ultimately expanded the market. The DeepSeek release in 2025 generated predictions of GPU demand deterioration, yet AI infrastructure investment accelerated. The transition from copper to optical networking, despite offering 10x bandwidth improvements, drove higher AI server capital outlays rather than cost reductions.
NAND Market Fundamentals Support Bullish Outlook
Rakesh highlighted that NAND memory capacity in AI servers has doubled during the previous twelve-month period. Spot market pricing has maintained an upward trajectory quarter after quarter.
He maintained that TurboQuant-type compression technologies would “enable larger large-language models, faster inference and better tokenomics, spurring more spending” throughout the artificial intelligence ecosystem.
Given both rising prices and robust fundamental demand dynamics, Mizuho projects that Micron and SanDisk could surpass consensus earnings forecasts in coming quarters.
SanDisk currently trades near $652, representing a discount to Mizuho’s $710 valuation target. Micron hovers around $347, below the firm’s $530 price objective.


