Back to home
Technology

Phison CEO claims NAND shortage could last a staggering 10 years — says memory 'supercycle' imminent and 'severe' 2026 shortages are at hand

Source

Tom's Hardware

Published

TL;DR

AI Generated

Phison Electronics Corporation's CEO, Pua Khein-Seng, predicts a NAND flash shortage starting in 2026 that could last for a decade, surpassing other industry estimates. Pua suggests a "supercycle" in memory demand is imminent, driven by the AI data center boom. As flash demand surges due to increased storage needs for AI models, companies like SanDisk, Micron, and Western Digital are freezing or raising prices. Pua anticipates SSDs will dominate over HDDs in data centers within the next decade, with flash memory remaining strong for years to come. Morgan Stanley also foresees a "memory supercycle" ahead, influenced by HBM demand and potential market shifts.

Read Full Article

Similar Articles

Microsoft announces 'world's most powerful' AI data center — 315-acre site to house 'hundreds of thousands' of Nvidia GPUs and enough fiber to circle the Earth 4.5 times

Microsoft announces 'world's most powerful' AI data center — 315-acre site to house 'hundreds of thousands' of Nvidia GPUs and enough fiber to circle the Earth 4.5 times

Microsoft has revealed plans for Fairwater, a massive AI data center set to become the "world's most powerful" upon its completion in early 2026. The facility, located in Mount Pleasant, Wisconsin, will span 315 acres and house "hundreds of thousands" of Nvidia GPUs connected by enough fiber to circle the Earth 4.5 times. Fairwater will utilize closed-loop water cooling to minimize water waste and will offer ten times the performance of today's fastest supercomputer. Microsoft is focused on sustainability, aiming to avoid increasing electricity costs for local communities. The construction involves impressive numbers, including 46.6 miles of deep foundation piles and storage systems equivalent to five football fields. Multiple similar data centers are also in progress across the United States.

Tom's Hardware
AI in Wyoming may soon use more electricity than state’s human residents

AI in Wyoming may soon use more electricity than state’s human residents

A new AI data center planned in Cheyenne, Wyoming, is set to consume more electricity than all homes in the state combined, with energy demands exceeding current residential, commercial, and industrial usage. The facility, a collaboration between Tallgrass and Crusoe, will start at 1.8 gigawatts and could scale up to 10 gigawatts of power usage annually. To meet these demands, the project will rely on its own gas generation and renewable energy sources due to the strain on the public grid. This initiative marks a significant shift for Wyoming, which currently exports a majority of its generated power to other states.

Ars Technica
OpenAI and partners are building a massive AI data center in Texas

OpenAI and partners are building a massive AI data center in Texas

OpenAI has partnered with Oracle to expand its Stargate AI infrastructure platform with 4.5 gigawatts of additional data center capacity in Texas, part of a reported $30 billion-per-year deal. The total Stargate capacity under development is set to exceed 5 gigawatts. The data center is located in Abilene, Texas, known for its evolving tech and manufacturing sectors, with existing tech employment opportunities from institutions like Dyess Air Force Base and universities. OpenAI's CEO, Sam Altman, expressed plans to expand Stargate's ambitions beyond the previously announced $500 billion commitment.

Ars Technica
Dr. L.C. Lu on TSMC Advanced Technology Design Solutions

Dr. L.C. Lu on TSMC Advanced Technology Design Solutions

Dr. L.C. Lu, a key figure at TSMC, focuses on design-technology co-optimization, packaging innovations, and AI-driven methodologies for next-gen semiconductor systems. TSMC emphasizes DTCO and DDCL innovations for scaling from N5 to A14 nodes, with NanoFlex and NanoFlex Pro architectures offering efficiency gains. N2P and N2U nodes incorporate advanced DTCO and power delivery optimizations, with hybrid dual-rail architectures achieving significant energy savings. TSMC collaborates with EDA partners for AI integration, enhancing productivity and design quality. Advanced packaging technologies like CoWoS and SoIC play a crucial role in enabling AI scaling, with memory bandwidth and interconnect performance scaling aggressively. TSMC addresses power delivery and thermal management challenges in AI systems through advanced solutions. TSMC's advancements in design methodologies and AI-driven automation promise improved productivity and scalability in chip-package co-design.

SemiWiki

We use cookies

We use cookies to ensure you get the best experience on our website. For more information on how we use cookies, please see our cookie policy.