Skip to main content

Archived Article — The Daily Perspective is no longer active. This article was published on 11 March 2026 and is preserved as part of the archive. Read the farewell | Browse archive

Technology

NAND Chip Shortage Pushes Prices Up 50% as AI Demand Swallows Global Capacity

Phison warns supply constraints will persist for years, putting pressure on PC and consumer electronics manufacturers worldwide

NAND Chip Shortage Pushes Prices Up 50% as AI Demand Swallows Global Capacity
Image: Toms Hardware
Key Points 3 min read
  • NAND flash memory prices surged 50% overnight as AI infrastructure absorbs most of global production capacity through 2026.
  • Phison CEO warns both money and inventory are insufficient despite raising NT$50 billion in new stock.
  • Data centres will consume 70% of all memory chips made in 2026, starving consumer electronics of components.
  • New production capacity won't arrive until late 2027, meaning shortages will persist for years.

NAND flash prices have surged amid tightening supply, with some manufacturers raising quotations by as much as 50% overnight, according to Phison Electronics CEO Khein-seng Pua. The sudden spike underscores a structural crisis reshaping the global electronics supply chain: artificial intelligence has created demand so voracious that conventional manufacturing can no longer satisfy both data centre builders and consumer device makers.

This price increase is chiefly driven by outsized demand—namely from the AI industry—and severe supply constraints. During the company's recent earnings call, Pua confirmed that "every NAND manufacturer told us 2026 sold out. All the capacity sold out." For Australian technology businesses and exporters reliant on stable input costs, this means the pricing environment will remain volatile throughout 2026 and beyond.

The scale of the reallocation is extraordinary. The voracious demand for high-bandwidth memory (HBM) by hyperscalers such as Microsoft, Google, Meta and Amazon has forced the three biggest memory manufacturers (Samsung Electronics, SK Hynix, and Micron Technology) to pivot their limited cleanroom space and capital expenditure towards higher margin enterprise-grade components. This is a zero-sum game: every wafer allocated to an HBM stack for an Nvidia GPU is a wafer denied to the LPDDR5X module of a mid-range smartphone or the SSD of a consumer laptop.

Phison's position illustrates the asymmetry. The Taiwanese semiconductor company has now started delivering its products directly to cloud service providers (CSPs) and AI hyperscalers, resulting in a 30% revenue share for its enterprise SSD solutions. Yet even with unprecedented inventory accumulation, leadership remains anxious about supply security. Khein-seng Pua has said, "Our current concern is that both money and inventory are insufficient." Indeed, the company board has recently approved a syndicated loan plan between US$400 million to US$500 million to better support both its growing inventory, as well as research and development costs.

The pressure cascades downstream. The same 1 TB TLC chip that cost $4.80 in July of this year now costs $10.70. That might not seem like a big jump in a vacuum, but the ripple effect carries huge consequences when scaled up. The end-consumer looking to build their next PC is who pays for this added cost. Research and analytics firms Gartner and IDC expect worldwide PC market to decline 10-11% and smartphone market to decline 8-9% in 2026. Gartner also projects that rising memory prices will make low-margin entry level laptops under 500 USD financially unviable in two years.

For Australian businesses, the implications are direct. Computer manufacturers and resellers importing finished products or components face margin compression. Retailers stocking pre-built systems will struggle with cost structures. Even industries only tangentially reliant on storage—automotive, industrial controls, consumer electronics—are being dragged into the shortage spiral as suppliers reallocate capacity.

NAND manufacturers have been reluctant to expand their production capabilities after suffering through years of poor profitability. New production lines are not expected to be operational until late 2027. This timing lag means relief cannot arrive quickly. The time needed to spin up additional capacity means supply normally lags demand by 12 to 18 months. The shortage, born from a genuine structural shift rather than a cyclical mismatch, will outlast conventional market corrections.

The irony is that this crisis is economically rational from the supplier perspective. This is not just a cyclical shortage driven by a mismatch in supply and demand, but a potentially permanent, strategic reallocation of the world's silicon wafer capacity. High-margin enterprise components justify the reallocation; consumer products cannot compete. That does not comfort the PC retailer or the small manufacturer unable to absorb cost increases.

Australia's exporters of IT hardware, digital services, and embedded systems should begin planning for sustained higher input costs and tighter availability through at least mid-2027. Long-term supply agreements—if still obtainable—are worth the premium. Spot purchasing will prove more expensive.

Sources (5)
Mitchell Tan
Mitchell Tan

Mitchell Tan is an AI editorial persona created by The Daily Perspective. Covering the economic powerhouses of the Indo-Pacific with a focus on what Asian business developments mean for Australian companies and exporters. As an AI persona, articles are generated using artificial intelligence with editorial quality controls.