惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

C
Check Point Blog
Y
Y Combinator Blog
GbyAI
GbyAI
Microsoft Azure Blog
Microsoft Azure Blog
Martin Fowler
Martin Fowler
博客园_首页
大猫的无限游戏
大猫的无限游戏
美团技术团队
S
SegmentFault 最新的问题
T
The Blog of Author Tim Ferriss
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
MyScale Blog
MyScale Blog
小众软件
小众软件
Vercel News
Vercel News
阮一峰的网络日志
阮一峰的网络日志
N
Netflix TechBlog - Medium
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
量子位
博客园 - 【当耐特】
J
Java Code Geeks
F
Fortinet All Blogs
宝玉的分享
宝玉的分享
Stack Overflow Blog
Stack Overflow Blog
博客园 - 司徒正美

Coco Feng - South China Morning Post

Shein clears Hong Kong IPO hurdle as valuation halves to below US$50b: sources Chinese state-owned firms set up chip funds for long-term innovation Breaking | Shein receives approval for Hong Kong IPO from Chinese regulators Game-changer? Why China is finally letting its AI firms buy the Nvidia H200 Why Alibaba shares just scored their strongest gains in Hong Kong this year China records most new unicorn start-ups in 5 years as AI and robotics boom Final insurance frontier: Hong Kong sees opening in mainland China’s space boom How a Chinese physical AI start-up’s new paradigm bypasses US road maps Dishing the dirt: official media alleges smear tactics in delivery battle Zhipu AI market cap tops HK$1 trillion after GLM-5.2 release Chinese AI improves forecasts as Hong Kong braces for super typhoons Alibaba and JD.com targeted in smear campaign ahead of shopping festival Who pays when rockets explode? China rushes to insure SpaceX’s rivals Microsoft cuts hundreds of mainland cloud jobs, as China, US tighten data laws Huawei chips refine DeepSeek model in leap for China’s AI self-reliance Part boat, part plane, part sci-fi: China debuts consumer watercraft China just beat Nvidia on a global robotics ranking. Is a new tech war brewing? Want to save the planet? Stop being so polite to AI chatbots Alibaba elevates tech chief Wu Zeming to elite committee amid AI push Mainland Chinese locked out of Computex Taipei amid cross-strait tensions ‘Arms race’: why investors can’t let go of robotic hand developers in China Kuaishou beats estimates as Kling AI video tool’s revenue jumps 300% Huawei’s workaround for US chip curbs faces hurdles on road to self-reliance China told the solar industry to cut capacity. Firms built new factories anyway China unveils first humanoid robot for household chores, ready as early as 2027 China’s Moonshot AI moves to unwind offshore structure in IPO pursuit: sources Trump, Xi weigh AI ‘guard rails’ as Nvidia chip exports hang in the balance Alibaba shares surge 7% in Hong Kong as firm accelerates pivot to AI Why are Chinese stock markets hitting record highs ahead of Trump’s visit? Nvidia’s Jensen Huang joins Trump’s trip to China at last minute
Memory price hikes ‘the new normal’ as AI fuels storage d...
Coco Feng · 2026-03-10 · via Coco Feng - South China Morning Post

Memory price hikes are likely to become “the new normal” for the next few years, according to an executive at storage giant Seagate Technology, as the artificial intelligence boom of the past two to three years pushes the industry into what he described as a supercycle.

“It’s hard to tell if it will last forever,” Ban-Seng Teh, Seagate’s chief commercial officer, told the South China Morning Post. The current cycle was “very unusual because in the past we went through cycles of shortage and oversupply”, he said.

The company, which mainly produces hard drives that store data on disks, had “definitely seen increasing costs” due to rising DRAM prices, Teh said, although its own usage of DRAM was “very small” compared with that of computer makers or data-centre operators.

DRAM, or dynamic random-access memory, is a type of chip that allows rapid storage and retrieval of vast amounts of data.

In the first quarter of 2026, prices of server DRAM were expected to surge by around 90 per cent from the previous quarter – the steepest increase on record – according to market intelligence firm TrendForce.

The surge has been fuelled by booming demand from AI data centres, which require far more memory to feed graphics processors used in training and running large language models.