Last week, Alphabet’s Google unveiled TurboQuant, a groundbreaking artificial intelligence (AI) algorithm that immediately sent shockwaves through the memory chip sector. The innovation, touted by researchers for its ability to reduce memory usage ‘by at least 6x and delivers up to 8x speedup, all with zero accuracy loss,’ effectively reducing the amount of memory needed by as much as 83%, triggered a sharp decline in the shares of leading memory chipmakers. Micron Technology (NASDAQ: MU) saw its stock fall 10%, while Sandisk Corporation (NASDAQ: SNDK) plunged 14% following the news, as investors feared a dramatic drop in demand for their semiconductors.
Google’s TurboQuant Redefines AI Efficiency
Google’s TurboQuant algorithm represents a significant leap forward in AI efficiency, particularly for large language models (LLMs). Its stated capability to drastically cut memory consumption and accelerate processing speeds without compromising accuracy has profound implications for the computational resources required to run advanced AI applications. The immediate market reaction reflected a conventional supply-demand concern: if AI systems need significantly less memory, then demand for memory chips, the core product of companies like Micron and Sandisk, would logically decrease, leading to reduced sales and revenue.
The Jevons Paradox: An Obscure Economic Counterpoint
However, some financial experts are urging caution against these immediate fears, pointing to an economic concept known as the Jevons paradox. First articulated by British economist William Stanley Jevons in his 1865 work, The Coal Question, this paradox suggests that increased efficiency in resource use, by reducing the cost of that resource, ultimately leads to an increase, rather than a decrease, in its overall consumption. Jevons observed this phenomenon with steam engines; as they became more efficient, the cost of coal decreased, which in turn prompted greater adoption and use of steam power, causing British coal consumption to triple between 1865 and 1900.
This historical pattern has been observed in various contexts. For instance, improvements in automobile fuel efficiency lowered the cost of driving per mile, encouraging consumers to drive more and, paradoxically, boosting overall fuel demand. The core principle is that efficiency gains make a resource more accessible and affordable, unlocking new applications and broader adoption that outweigh the per-unit reduction in usage.
Applying the Paradox to AI Memory Demand
The Jevons paradox offers a compelling counter-narrative to the initial investor panic surrounding Google’s TurboQuant. While the algorithm makes running LLMs more efficient and could reduce the per-unit memory requirement, it is also expected to lower the effective cost of deploying and operating AI. This reduction in cost is anticipated to fuel a much wider adoption of AI technologies across various industries and applications. Consequently, the overall demand for memory chips, despite their individual efficiency, could surge as AI becomes more pervasive and integrated into new systems and services.
Analyst Insights and a Potential Buying Opportunity
This perspective is echoed by market analysts. Just this week, Mizuho analyst Vijay Rakesh reiterated his ‘outperform (buy)’ ratings on both Micron and Sandisk. Rakesh posited that advancements like TurboQuant are fundamentally positive for the memory sector. He explicitly cited the Jevons paradox, stating that ‘performance improvements will drive further adoption of AI and strengthen demand for key components such as memory chips.’ Rakesh further elaborated in a note to clients that TurboQuant ‘will enable larger [LLMs], faster inference and better tokenomics, spurring more spending.’
This suggests that the initial pullback in Micron and Sandisk stocks, driven by fears of declining memory sales, might represent a strategic buying opportunity for investors who understand the historical parallels of the Jevons paradox.
Micron and Sandisk: Strong Financials Amidst Market Volatility
Beyond the economic theory, both Micron and Sandisk exhibit robust financial indicators that support a bullish outlook. Micron stock has demonstrated significant growth, gaining more than 500% over the past three years as of April 4, 2026. Despite this impressive run, the company’s valuation remains attractive, selling for just 17 times earnings and boasting a price/earnings-to-growth (PEG) ratio of 0.04, well below the standard threshold of 1 for an undervalued stock.
Micron’s management provided a compelling Q3 outlook, forecasting revenue of $33.5 billion, which would represent a remarkable 260% year-over-year growth and 40% quarter-over-quarter expansion. The company also projects its gross margin to increase by 660 basis points, from 74.4% to approximately 81%, pushing adjusted diluted earnings per share to roughly $19.15, a tenfold increase.
Sandisk, which was spun off from Western Digital in February 2025, has experienced an even more dramatic surge, with its stock price climbing 1,850% since its separation. Similar to Micron, Sandisk trades at a modest 15 times earnings with an exceptionally low PEG ratio of 0.01. For its upcoming third quarter, Sandisk’s forecast calls for revenue of $4.6 billion at the midpoint of its guidance, indicating 171% growth. Management anticipates a gross margin of 65.9% at the midpoint, nearly tripling last year’s 22.5%.
While these growth targets are ambitious, and the deployment of TurboQuant could indeed impact the per-unit price of memory chips, historical evidence, particularly the Jevons paradox, strongly suggests that efficiency gains will be channeled into broader AI adoption, ultimately fueling even greater aggregate demand for memory. The current valuations of Micron and Sandisk, with little growth seemingly ‘baked in’ at their present prices, indicate that these companies could represent a compelling investment opportunity as the AI revolution continues to unfold.
