--- title: "Morgan Stanley Comments on NVIDIA 'Memory Halving' Rumors: This Is a Real Shortage, Supply Constraints Have Become the Biggest Obstacle to AI" type: "News" locale: "en" url: "https://longbridge.com/en/news/289047183.md" description: "Morgan Stanley believes that NVIDIA's reduction of Vera Rubin rack memory configurations is not a sign of cooling AI demand, but rather a reflection of intensifying DRAM supply shortages. Industry data from April showed that DRAM and NAND sales and price growth hit record highs, with memory remaining the biggest bottleneck in AI infrastructure. Based on continued tight supply and demand, Morgan Stanley significantly raised its revenue forecasts for the semiconductor industry and remains bullish on leaders in the AI and memory supply chains such as Micron Tech, Sandisk, and NVIDIA" datetime: "2026-06-08T11:22:45.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/289047183.md) - [en](https://longbridge.com/en/news/289047183.md) - [zh-HK](https://longbridge.com/zh-HK/news/289047183.md) --- # Morgan Stanley Comments on NVIDIA 'Memory Halving' Rumors: This Is a Real Shortage, Supply Constraints Have Become the Biggest Obstacle to AI Rumors that NVIDIA is cutting memory configurations for its Vera Rubin racks have shaken the market, but Morgan Stanley argues that this precisely confirms the severity of the memory shortage, rather than signaling a cooldown in AI demand. Last week, research firm SemiAnalysis released a report stating that NVIDIA plans to reduce the LPDDR5 memory capacity of Vera Rubin racks from 55 TB to 28 TB, and shrink SOCAMM memory module specifications from 192 GB to 96 GB. The market interpreted this news as a sign of weakening AI demand, directly causing Micron Tech (MU) shares to plunge 13% on the same day it received NVIDIA HBM4 certification, marking its largest single-day drop since April 2025. Morgan Stanley semiconductor analyst Joseph Moore promptly released a research report refuting this interpretation. According to Zhui Feng Trading Desk, Morgan Stanley explicitly stated in its latest report that it has verified that some racks will indeed be shipped with lower configurations, but emphasized that this adjustment is entirely due to supply-side constraints, not weak demand. Meanwhile, the bank raised its 2026 global semiconductor industry revenue forecast from $807 billion to $880 billion and maintained its Overweight ratings on MU and SNDK, with target prices of $1,050 and $1,750, respectively. ## Rumors Are True, But Logic Was Misinterpreted by the Market Morgan Stanley confirmed in its report that NVIDIA is indeed reducing memory configurations for Vera Rubin racks, but emphasized that the market's interpretation of the event is completely off base. The report pointed out that NVIDIA and cloud computing customers are willing to buy every gigabyte of SOCAMM memory they can get, and will immediately revert to higher configurations once supply catches up. Morgan Stanley believes that the sole purpose of this adjustment is to minimize the impact of the DRAM shortage on GPU rack sales, which in itself is evidence of a real shortage—rather than a product of the reorder window period. In terms of scale impact, assuming 53,000 to 70,000 racks are built next year, rack SOCAMM demand under the 55 TB configuration would approach 5% of total global DRAM demand. If halved across the board, the most extreme scenario would reduce demand by approximately 1.4 million TB, equivalent to more than 2% of the $6.2 million TB market size, affecting the higher-value segments of the market. Morgan Stanley expects that higher configurations will return to supply shortly after shipments begin. ## April SIA Data: No Quick Fix for Memory Shortage Morgan Stanley also commented on the April Semiconductor Industry Association (SIA) billing data released on June 5, believing that this further reinforces the judgment that memory supply remains tightly constrained. Overall semiconductor sales in April decreased by 2.2% month-over-month, significantly better than Morgan Stanley's estimate of -12.1% and the 10-year seasonal average of -10.6%, with year-over-year growth accelerating to 106.4%. Among these, memory performance was particularly outstanding: DRAM sales decreased by only 3.7% month-over-month, far better than the estimated -24.2% and the 5-year average of -27.6%, with a year-over-year increase of 375.3%. The 3-month rolling year-over-year growth rate reached 298.5%, hitting a record high since 2001, with average selling price year-over-year growth accelerating for nine consecutive quarters; NAND sales decreased by 4.2% month-over-month, also significantly better than the estimated -11.2%, with a year-over-year increase of 366.0%. The 3-month rolling year-over-year revenue growth rate of 307.0% and average selling price growth rate of 213.4% both set historical records. Morgan Stanley pointed out that the April data confirmed the judgment that there is no quick solution to the memory shortage. DRAM has become the main bottleneck in AI infrastructure construction. Intensive HBM wafer capacity, cleanroom and EUV equipment constraints, along with limited incremental NAND capacity, jointly support a "higher for longer" pricing environment. The bank believes that long-term agreements (LTAs) are a symptom of supply tightness and hyperscale cloud providers competing for capacity, rather than the cause of this cycle. ## Raising Forecasts, Bullish on Memory and AI Supply Chain Based on the above judgments, Morgan Stanley significantly raised its industry forecasts. The 2026 global semiconductor revenue forecast was raised from $910 billion to $1.607 trillion, with the expected year-over-year growth rate increasing from +91% to +103%, mainly driven by memory. Specifically, the DRAM forecast jumped from $150.3 billion to $581.9 billion, and NAND rose from $67.7 billion to $293.5 billion. For 2027, the bank expects industry revenue to grow further by 22% to $1.96 trillion, primarily driven by the continued pass-through of memory pricing. At the individual stock level, Morgan Stanley maintains its Overweight rating on MU, with a base target price of $1,050 (current price $864, potential upside of 22%); it maintains its Overweight rating on SNDK, with a base target price of $1,750 (current price $1,559, potential upside of 12%). In addition, the bank remains bullish on AI infrastructure beneficiaries NVDA (Overweight, target price $288), AVGO (Overweight, target price $502), and ALAB (Overweight, target price $240), as well as capital equipment/supply chain targets LRCX, KLAC, and MKSI. ## Broad Market Improvement Synchronizes, Stronger Signals of Cycle Diffusion Morgan Stanley pointed out that the highlights of the April SIA data were not limited to memory; the broad-based market's comprehensive outperformance is also worth noting, indicating that the current semiconductor upcycle is spreading from being AI-driven to broader fields. By category, discrete devices, analog chips, MCUs, and MPUs all exceeded Morgan Stanley's estimates and the 10-year seasonal average. The industrial sector is re-accelerating from cyclical lows, with demand expanding into non-AI-related areas and inventory digestion continuing to progress. The bank stated that although supply chains have seen price increases, the main driver is cost pass-through rather than value recapture, with few exceptions such as ADI. Morgan Stanley concluded that the April data further proves that the current semiconductor upcycle is no longer just a narrow AI-driven rally, but is increasingly evolving into a broader supply-constrained upcycle. In this context, the bank maintains a positive view on analog/MCU suppliers ADI, NXP, and ALGM. ### Related Stocks - [NVDA.US](https://longbridge.com/en/quote/NVDA.US.md) - [SNDK.US](https://longbridge.com/en/quote/SNDK.US.md) - [MU.US](https://longbridge.com/en/quote/MU.US.md) - [MUYY.US](https://longbridge.com/en/quote/MUYY.US.md) - [NVDW.US](https://longbridge.com/en/quote/NVDW.US.md) - [FTXL.US](https://longbridge.com/en/quote/FTXL.US.md) - [PSI.US](https://longbridge.com/en/quote/PSI.US.md) - [XSD.US](https://longbridge.com/en/quote/XSD.US.md) - [SOXQ.US](https://longbridge.com/en/quote/SOXQ.US.md) - [SOXL.US](https://longbridge.com/en/quote/SOXL.US.md) - [SOXX.US](https://longbridge.com/en/quote/SOXX.US.md) - [NVDO.US](https://longbridge.com/en/quote/NVDO.US.md) - [SMH.US](https://longbridge.com/en/quote/SMH.US.md) - [SNXX.US](https://longbridge.com/en/quote/SNXX.US.md) - [SNDU.US](https://longbridge.com/en/quote/SNDU.US.md) - [SNDQ.US](https://longbridge.com/en/quote/SNDQ.US.md) - [SNDG.US](https://longbridge.com/en/quote/SNDG.US.md) - [NYYY.US](https://longbridge.com/en/quote/NYYY.US.md) - [NVDB.US](https://longbridge.com/en/quote/NVDB.US.md) - [NVYY.US](https://longbridge.com/en/quote/NVYY.US.md) - [NVDG.US](https://longbridge.com/en/quote/NVDG.US.md) - [MULL.US](https://longbridge.com/en/quote/MULL.US.md) - [NVDQ.US](https://longbridge.com/en/quote/NVDQ.US.md) - [MUU.US](https://longbridge.com/en/quote/MUU.US.md) - [DRAM.US](https://longbridge.com/en/quote/DRAM.US.md) - [NVDL.US](https://longbridge.com/en/quote/NVDL.US.md) - [NVDU.US](https://longbridge.com/en/quote/NVDU.US.md) - [MUD.US](https://longbridge.com/en/quote/MUD.US.md) - [NVDS.US](https://longbridge.com/en/quote/NVDS.US.md) - [NVDY.US](https://longbridge.com/en/quote/NVDY.US.md) - [NVDX.US](https://longbridge.com/en/quote/NVDX.US.md) - [NVDD.US](https://longbridge.com/en/quote/NVDD.US.md) - [NVD.US](https://longbridge.com/en/quote/NVD.US.md) - [07788.HK](https://longbridge.com/en/quote/07788.HK.md) - [07388.HK](https://longbridge.com/en/quote/07388.HK.md) - [HBMX.US](https://longbridge.com/en/quote/HBMX.US.md) - [MS.US](https://longbridge.com/en/quote/MS.US.md) - [AVGO.US](https://longbridge.com/en/quote/AVGO.US.md) - [LRCX.US](https://longbridge.com/en/quote/LRCX.US.md) - [KLAC.US](https://longbridge.com/en/quote/KLAC.US.md) - [MKSI.US](https://longbridge.com/en/quote/MKSI.US.md) - [ADI.US](https://longbridge.com/en/quote/ADI.US.md) - [NXPI.US](https://longbridge.com/en/quote/NXPI.US.md) - [ALGM.US](https://longbridge.com/en/quote/ALGM.US.md) - [MS-O.US](https://longbridge.com/en/quote/MS-O.US.md) - [MS-Q.US](https://longbridge.com/en/quote/MS-Q.US.md) - [MS-E.US](https://longbridge.com/en/quote/MS-E.US.md) - [MS-I.US](https://longbridge.com/en/quote/MS-I.US.md) - [MS-L.US](https://longbridge.com/en/quote/MS-L.US.md) - [MS-P.US](https://longbridge.com/en/quote/MS-P.US.md) - [MS-A.US](https://longbridge.com/en/quote/MS-A.US.md) - [MS-F.US](https://longbridge.com/en/quote/MS-F.US.md) - [MS-K.US](https://longbridge.com/en/quote/MS-K.US.md) - [NVD.DE](https://longbridge.com/en/quote/NVD.DE.md) - [SNDKV.US](https://longbridge.com/en/quote/SNDKV.US.md) ## Related News & Research - [ANALYSIS-Nvidia's AI PC push banks on unproven demand beyond niche users](https://longbridge.com/en/news/289042406.md) - [Here's why the VanEck Semiconductor ETF soared in May and is a great way to play AI spending](https://longbridge.com/en/news/288596388.md) - [Better Buy: Sandisk or Nvidia Stock?](https://longbridge.com/en/news/288986978.md) - [Why Sandisk Stock Is the AI Storage Supercycle’s Biggest Winner](https://longbridge.com/en/news/288587191.md) - [South Korea's Naver to build gigawatt-scale AI factories using Nvidia technology](https://longbridge.com/en/news/288973405.md)