Key Takeaways
- Memory chip stocks experienced approximately 10% decline last month, with Micron and Sandisk dropping more than 10% following Google’s TurboQuant announcement
- Google’s TurboQuant technology promises to cut AI memory requirements by as much as 6x, triggering investor concerns
- Morgan Stanley characterizes the recent decline as a healthy correction rather than a fundamental problem
- Memory capacity has emerged as the primary constraint in AI infrastructure expansion, surpassing GPU availability
- Morgan Stanley maintains Overweight ratings with $520 target for Micron and $690 for Sandisk
Morgan Stanley continues to support memory semiconductor manufacturers following a significant market downturn that shook investor confidence in late March.
The iShares Semiconductor ETF experienced approximately 10% losses during the past month. Multiple factors contributed to the decline, including valuation concerns, demand uncertainty, and emerging AI innovations.
Google unveiled a novel compression technology named TurboQuant on March 24. This innovation purportedly cuts memory requirements for operating AI models by as much as six times. The announcement triggered alarm among market participants.
Both Micron and Sandisk experienced declines exceeding 10% in the trading sessions after the announcement. Micron settled at $357 on March 27, maintaining a 25% gain for the year-to-date period.
Joseph Moore, analyst at Morgan Stanley, challenged the negative market reaction in a research report distributed on March 26.
Moore confirmed Overweight ratings for both Micron and Sandisk. The firm’s price targets stay unchanged at $520 and $690, respectively.
According to Moore, the selloff represents “a healthy pricing in of durability concerns” instead of an actual change in underlying demand dynamics. The investment bank argues that memory manufacturers’ fundamental strength is “more durable than the market thinks.”
Memory Capacity Replaces GPUs as Primary AI Infrastructure Constraint
Throughout the previous two years, Nvidia’s graphics processing units dominated headlines as the critical component driving AI infrastructure investments. While this remains accurate, Morgan Stanley contends that memory has evolved into the constraining element.
“Memory is a bottleneck, increasingly the bottleneck, to AI builds,” the research team stated. They highlighted that clients are now making advance payments for high-volume orders, demonstrating how constrained supply has become.
According to Moore, DRAM availability has completely evaporated. “Everywhere we look we see indications that it is a true bottleneck,” he noted.
The investment bank projects AI’s portion of semiconductor expenditure could reach “well north of 50%.” Expanding production capacity is unlikely to match such robust demand levels.
Morgan Stanley’s Assessment of TurboQuant Technology
Morgan Stanley specifically analyzed Google’s TurboQuant release, arguing that investors misinterpreted its implications.
The compression technology exclusively targets KV Cache memory, not total memory consumption. “They are just talking about KV Cache memory, not memory overall,” the analysts clarified.
KV Cache utilizes high-bandwidth memory, a specialized and scarce variety. Morgan Stanley characterized TurboQuant as “normal course productivity improvement,” rather than a demand-threatening breakthrough.
The financial institution doesn’t anticipate gross margins approaching 81% to persist indefinitely. However, analysts see minimal justification for near-term margin compression.
Morgan Stanley also emphasized robust prospects for free cash flow generation among memory manufacturers. The firm determined that “duration is all that matters,” and by that metric, signals “all appear positive.”
Micron and Sandisk retained their Overweight ratings as of March 26, 2026.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
