Micron Invests $24 Billion in Singapore Factory, But Your RAM Sticks Are Not in the Plan

marsbitPublicado a 2026-01-27Actualizado a 2026-01-27

Resumen

Micron announces a $24 billion investment to build an advanced NAND wafer fabrication plant in Singapore, with production starting in late 2028. This follows a previous $7 billion investment in a high-bandwidth memory (HBM) advanced packaging facility at the same site, bringing total new investments in Singapore to over $30 billion. The move comes as Micron shifts its focus toward AI-driven demand and away from consumer products, having recently discontinued its Crucial consumer brand, which sold retail memory modules and SSDs. The strategic pivot is driven by surging demand from AI and data centers. Micron’s HBM revenue reached $1.98 billion in Q3 FY2025, with full-year 2026 HBM supply already sold out. The company predicts the HBM market will grow to around $100 billion by 2028, surpassing the entire DRAM market of 2024. In contrast, consumer memory market is lower-margin and highly competitive. This supply shift has contributed to significant price increases in consumer memory products, with DDR4 and DDR5 spot prices rising over 150% and 300% respectively in 2025. Although Samsung and SK Hynix have extended DDR4 production in response to soaring prices, overall supply remains tight. As AI continues to drive memory demand, consumer memory costs are expected to stay elevated.

Author: David, Shenchao TechFlow

At this time last year, a 16GB DDR4 RAM stick cost just over 200 RMB. Now, the same model is priced at 600 RMB, with some models even approaching 700 RMB.

The direct reason for the price increase is a tightening of supply. The three major memory giants have collectively shifted their production capacity towards the AI market. And Micron is the latest to lay its cards on the table.

On January 27, Micron Technology announced a $24 billion (approximately S$31 billion) investment to build an advanced NAND (flash memory) wafer fabrication plant in Singapore. The investment will be phased over 10 years, with shipments starting in the second half of 2028. Singapore's Deputy Prime Minister Gan Kim Yong attended the groundbreaking ceremony.

This is Micron's second major investment in Singapore.

In January 2025, Micron had already begun construction on a $7 billion HBM (High Bandwidth Memory) advanced packaging facility within the same campus, scheduled for production in 2027. Combined, these two projects represent over $30 billion in new investments by Micron in Singapore.

Image source: Lianhe Zaobao, photo by Tang Jiahong

Yet, this is the same company that just two months ago announced the shutdown of its 29-year-old Crucial consumer brand—the memory sticks and solid-state drives you can buy on JD.com and Taobao.

On one hand, spending billions to build factories; on the other, cutting the consumer business. Both point to the same clue: the demand for memory from AI remains high.

Capacity Reallocation After China Market Contraction

Micron's increased investment in Singapore has a clear geopolitical background.

In May 2023, China's Cyberspace Administration announced that Micron's products had not passed the cybersecurity review and required domestic critical information infrastructure operators to stop procurement. In 2018, the Chinese mainland market contributed 58% of Micron's revenue (approximately $17.36 billion); by fiscal year 2022, this had fallen to 10.8% (approximately $3.31 billion).

Production capacity needs a new outlet.

Currently, Micron already has three 3D NAND fabrication plants and several packaging and testing facilities in Singapore, employing about 9,000 people and producing 98% of Micron's flash memory chips.

Micron stated in its latest announcement that the HBM factory is expected to make a significant contribution to supply by 2027 and mentioned that as HBM becomes part of the Singapore manufacturing operations, "synergistic effects" in both NAND and DRAM (Dynamic Random-Access Memory, the common memory chips) production are anticipated.

But the priority of this "synergy" is clear: HBM capacity for 2026 is completely sold out, data center customer demand cannot be fully met, and the consumer brand is being shut down simultaneously.

Within the same campus, consumer-grade production lines are making way for AI production lines.

The Business Logic of Exiting the Consumer Market

On December 3, 2025, Micron announced via its official website a complete exit from the Crucial consumer business, including the sales of RAM sticks and SSD (Solid State Drive) products through global retailers, e-commerce platforms, and distributors.

Shipments will continue until the end of February 2026, at which point this brand, founded in 1996, will bid farewell to the retail market.

In the statement, Micron's Executive Vice President, Sadana, stated that the surge in memory and storage demand driven by AI in data centers prompted Micron to make the "difficult decision" to exit the Crucial consumer business in order to more effectively supply and support large strategic customers in high-growth areas.

Data supports this judgment.

Micron's fiscal year 2025 third-quarter HBM product income reached $1.98 billion, with an annualized revenue close to $8 billion. According to TrendForce data, the unit price of HBM memory required for AI servers is about 8 times that of ordinary servers.

Micron predicts the HBM market size will grow from approximately $35 billion in 2025 to about $100 billion in 2028, by which time it will exceed the size of the entire DRAM market in 2024.

In contrast, consumer-grade storage is a business with thin profits and fierce competition.

Micron's exit means the world's major consumer DRAM manufacturers are reduced from three—Samsung, SK Hynix, Micron—to two.

The AI Rally of Memory Stocks

The capital market has already priced in this shift.

According to statistics from *National Business Daily*, memory stocks occupied the top four spots on the U.S. stock gainers list for the full year 2025:

SanDisk rose 577%, becoming the S&P 500's top performer for the year; Western Digital rose 281%; Micron rose 236%; Seagate rose 216%.

For comparison, Nvidia rose 39% in the same period, ranking 71st.

It was reported that tech giants like Google and Amazon presented Micron with "unlimited purchase demands" in October 2025. Micron CEO Mehrotra revealed on an earnings call that the company's entire HBM supply for 2026 is already sold out, and currently, it can only meet half to two-thirds of the demand from key customers.

Micron's latest financial report shows revenue of $13.6 billion for September-November 2025, a year-on-year increase of 57%, with DRAM revenue of $10.8 billion, up 69% year-on-year.

Consumer Side: Price Hikes May Continue

Samsung, SK Hynix, and Micron had all previously announced plans to gradually phase out DDR4 production from late 2025 to early 2026. Micron issued an end-of-life notice for DDR4/LPDDR4 in June.

Although Samsung and SK Hynix later extended their DDR4 production plans to the end of 2026 due to soaring DDR4 prices, the overall trend of tightening supply remains unchanged.

The price reaction has been intense. According to TrendForce data, since 2025, the spot price of DDR5 memory chips has risen over 300%, while DDR4 has risen over 150%. DDR4 has even seen a "price inversion," with the price of some specifications exceeding that of DDR5.

ADATA Chairman Chen Libai also publicly stated that the comprehensive shortage and price increase across the four major storage categories—DRAM, NAND, SSD, HDD—is a situation he is witnessing for the first time in over thirty years in the industry.

If you want to add memory to your old computer or build a new desktop, the cost remains high. Because AI needs computing power, and computing power needs memory. Your old computer is probably still behind AI demand in the queue.

Memory stocks have risen, and memory stick prices have risen too.

One is an investment opportunity; the other is paying the bill for that opportunity.

Preguntas relacionadas

QWhy did Micron announce the closure of its Crucial consumer brand?

AMicron closed its Crucial consumer brand to reallocate resources and supply to meet the surging demand for memory and storage from data centers and large strategic customers driven by AI, as the consumer market is less profitable and highly competitive.

QWhat is the total investment Micron is making in Singapore for new manufacturing facilities?

AMicron is investing a total of over $30 billion in Singapore, including $24 billion for a new NAND wafer fab and over $7 billion for an HBM advanced packaging plant.

QHow has the price of consumer memory modules changed recently according to the article?

AThe price of a 16GB DDR4 memory module has increased from over 200 yuan to about 600 yuan, with some models nearing 700 yuan, due to tightened supply and shifted production focus to AI.

QWhat percentage of Micron's revenue came from the Chinese market in 2022 compared to 2018?

AIn 2018, the Chinese market contributed 58% of Micron's revenue, but by the 2022 fiscal year, it had decreased to 10.8%.

QWhich storage-related stocks were the top performers in the U.S. stock market in 2025 according to the article?

AIn 2025, the top-performing U.S. storage stocks were SanDisk (up 577%), Western Digital (up 281%), Micron (up 236%), and Seagate (up 216%).

Lecturas Relacionadas

Agent Race Ends, Super Workbench Takes Over

The era of fragmented AI agents is ending. Over the past month, China's tech giants—Tencent, Alibaba, and ByteDance—have simultaneously shifted strategy: instead of launching new, standalone AI agents, they are consolidating their various agent projects into unified "super workbenches." Tencent integrated its QClaw teams into WorkBuddy, a strategic product hailed as a potential third flagship after QQ and WeChat. Alibaba is merging its QoderWork, Wukong, and MuleRun agents into a new "Qianwen Office" platform under DingTalk's leadership. ByteDance rebranded its TRAE SOLO coding agent to TRAE Work, signaling a broader focus on workflow collaboration. This convergence marks a pivotal industry consensus. The initial exploration phase, where companies rapidly built numerous overlapping agents for different scenarios, proved costly and inefficient. With open-source tools eroding technical barriers, competition has shifted from agent creation to resource consolidation and cost control. Historically, platform wars are won not by creating more products, but by simplifying them—as seen with browsers unifying web access and super-apps consolidating services. Now, the "super workbench" aims to become the unified AI entry point for work. This reflects a deeper market realization: the primary audience for AI is no longer just programmers (a market in the tens of millions) but all knowledge workers (a market of billions). The real opportunity lies in augmenting everyday tasks—managing emails, documents, data, and meetings—across the entire workday. The core battleground is becoming control over the primary AI entry point that employees use daily. Tencent's WorkBuddy leverages WeChat and Tencent Docs; Alibaba's Qianwen Office taps into DingTalk's organizational data; ByteDance's TRAE Work integrates with Feishu's workflows. Whoever owns this "super workbench" gains strategic control over orchestrating enterprise data and APIs. This shift is redefining enterprise software. Traditional SaaS applications, valued for their user interfaces, will recede into the background. Their core functionalities will be exposed as standardized "Skills" or APIs for the super workbench's agents to invoke. Software value will shift from selling user seats to charging based on API calls and outcomes delivered. The evolution of agents is moving through clear stages: first as novel standalone products, then as consolidated primary work entry points, and finally as pervasive, invisible capabilities embedded into the digital fabric. The recent moves by major tech firms signal the transition from the first stage into the second, accelerating toward the third. In the end, the most successful agent technology may become invisible—like electricity or the HTTP protocol—a fundamental, unnamed infrastructure powering work itself.

marsbitHace 18 min(s)

Agent Race Ends, Super Workbench Takes Over

marsbitHace 18 min(s)

Michael Saylor: 110 Reasons to Oppose BIP-110

Michael Saylor presents 110 arguments against Bitcoin Improvement Proposal (BIP) 110, a soft fork aimed at restricting certain non-monetary data storage uses (like inscriptions) on the Bitcoin blockchain. He acknowledges the proponents' valid concerns—such as node costs, fee pressure, and preserving Bitcoin's monetary focus—but fundamentally disagrees with the proposed solution. Saylor argues that BIP 110 represents a dangerous precedent of using consensus rules to enforce value judgments on transaction validity, moving away from Bitcoin's core principles of neutrality and permissionless innovation. His key objections are organized into eleven categories: 1) It violates neutrality and hard consensus by banning currently valid transactions. 2) It fails to meet the high burden of proof required for a consensus change, lacking concrete data on the alleged crisis. 3) Its seven bundled technical restrictions are overly broad, targeting generic script functionalities and blocking future upgrade paths. 4) It sacrifices compatibility and future optionality by closing off designed upgrade hooks. 5) Its temporary rules add significant complexity (grandfathering, expiry states) without sufficient justification. 6) The economic and security impacts, particularly on miner revenue and fee markets, are uncertain and unmodeled. 7) Superior, market-based tools (fee markets, relay/mining policies) already exist to manage blockchain load. 8) It stifles innovation by creating a chilling effect for developers. 9) Its modified activation mechanism (55% threshold, forced signaling) is aggressive and risks network splits. 10) The precedent it sets—using consensus to suppress disliked but legal uses—is more dangerous than the problem it aims to solve. 11) A better path exists: improving measurements, refining resource-based policies, and allowing market forces to work. Saylor concludes that Bitcoin's strength lies in its neutral rules, open markets, and hard consensus. Changing these foundational elements to target specific use cases is an unnecessary and risky "iatrogenic" intervention. He advocates for guarding Bitcoin's neutrality rather than acting as its redeemer.

marsbitHace 34 min(s)

Michael Saylor: 110 Reasons to Oppose BIP-110

marsbitHace 34 min(s)

United Stables Crosses $1B As Chainlink Data Feeds Secure U Token Collateral

United Stables' U token has surpassed $1 billion in market capitalization, with Chainlink Data Feeds providing the critical pricing and collateral data infrastructure across its deployment chains. This milestone highlights that robust oracle infrastructure is essential for stablecoin credibility, as DeFi protocols require reliable data for integration. Chainlink's role is foundational, enabling automated collateral auditing and pricing verification, which helps make the U token safer for the broader market to evaluate and use. While this strengthens Chainlink's position in the stablecoin infrastructure narrative, it does not directly guarantee increased fees or value for LINK token holders. The growth of U indicates both widening competition in the stablecoin sector and the increasing dependence of scalable stablecoins on dependable external data services.

bitcoinistHace 1 hora(s)

United Stables Crosses $1B As Chainlink Data Feeds Secure U Token Collateral

bitcoinistHace 1 hora(s)

BonkDAO Treasury Drain Shows Solana Governance Risk Is Real

The BonkDAO treasury suffered an approximately $20 million drain due to a malicious governance proposal passed through the Realms platform on Solana. This incident highlights governance as a critical attack surface for DAOs, distinct from smart contract exploits. The attack leveraged the DAO's own voting mechanics to move funds, demonstrating that flawed governance design—such as weak proposal rules or voter weight calculations—can be exploited even when the underlying blockchain functions correctly. The event serves as a warning for Solana DAOs to rigorously review and strengthen their governance safeguards, including quorum thresholds, timelocks, and execution permissions. While damaging to community trust, especially for the prominent BONK meme ecosystem, the issue is not a failure of the Solana base layer but a widespread application-level risk common across blockchain ecosystems.

bitcoinistHace 2 hora(s)

BonkDAO Treasury Drain Shows Solana Governance Risk Is Real

bitcoinistHace 2 hora(s)

Midnight Token Falls After $13M Wanchain Bridge Exploit

Midnight's NIGHT token plummeted roughly 30% following a $13.2 million exploit on a Wanchain cross-chain bridge. The attacker drained 515 million NIGHT tokens by exploiting a signature reuse flaw in the bridge's infrastructure, not the core Cardano or Midnight networks. This incident highlights the persistent vulnerability of cross-chain bridges, which are critical for asset transfers but introduce complex security risks. While the underlying blockchains remained secure, the token's price suffered due to liquidity disruption and lost market confidence. The response from Wanchain, including pausing the affected bridge and providing clear communication on recovery efforts, will be crucial for stabilizing the situation. The event serves as a reminder for traders to factor in bridge-related risks when evaluating tokens reliant on cross-chain liquidity.

bitcoinistHace 2 hora(s)

Midnight Token Falls After $13M Wanchain Bridge Exploit

bitcoinistHace 2 hora(s)

Trading

Spot
活动图片