Technology News
Latest technology news and updates
Total: 301
SK Hynix Weighs U.S. AI Investment Hub to Centralise Group’s AI Bets
SK Hynix is exploring the creation of an AI investment arm in the United States to centralise group-level AI investments and speed strategic decisions. The plan would place the company closer to cloud customers, AI start‑ups and investors, and could shift SK Hynix toward higher‑value participation in the AI infrastructure ecosystem.
Retail Demo Backfires: Xiaomi Staff Tosses Phone to Prove Durability — Screen Cracks, Public Questions Follow
A Xiaomi store employee’s attempt to demonstrate a phone’s drop resistance ended with the device’s screen cracking, and the clip sparked online mockery and questions about who should pay for repairs. The incident underscores how live retail demonstrations can backfire, creating reputational and liability headaches for manufacturers and retailers.
Ma Huateng’s AI U‑Turn: From Caution to a High‑Stakes Bet on WeChat and Compute
Over three years Tencent’s stance on AI has shifted from cautious integration to active, capital‑heavy deployment. Ma Huateng has reorganised AI teams, beefed up compute spending and launched a CNY 1bn promotional push to seed an AI assistant inside WeChat, betting the company’s social graph and ad franchise can translate into AI dominance.
Robots Invade China’s Spring Festival Gala: A High‑Stakes Pre‑IPO Roadshow
China’s top New Year variety show has become a critical marketing stage for domestic robot start‑ups seeking valuation uplift and orders before IPOs. Three robot companies — Unitree, Magic Atom and Galaxy General — have secured gala roles, prompting industry debate over whether polished performances obscure the tougher work of scaling and commercialization.
China’s Big Tech Turns Spring Festival Red Envelopes into an AI Battleground
Tencent’s recent 1 billion yuan Spring Festival giveaway has escalated a familiar marketing ritual into a proxy arena for China’s AI race. Big tech firms are shifting from model-centric competition to full-stack battles that combine massive capital expenditure, distributional advantage and ecosystem playbooks, with 2026 poised as a potential inflection point for AI monetisation.
Memory Makers Ride an AI-Fuelled Supercycle as Prices Soar — and Few Can Stop It
A surge in demand for AI‑related storage is driving rapid price rises across DRAM and NAND, with suppliers shifting to flexible, quarterly pricing and prioritising high‑margin AI products. Limited capacity growth — because investment is being spent on process upgrades rather than volume expansion — means the shortage looks structural and could persist through 2026–27, benefiting memory vendors but squeezing OEMs and raising the cost of scaling AI services.
Tencent's Sogou Keyboard Goes All‑In on AI, Adding Multi‑Language Real‑Time Translation
Tencent has launched Sogou Input Method 20.0, upgrading voice, typing and translation with model‑level AI and adding a 'HunYuan champion' translation model that supports over 30 languages. The change deepens Tencent's control over a key user interface and raises questions about data use, privacy and the company’s competitive posture in the AI arms race.
EU Designates WhatsApp a ‘Very Large’ Platform, Pushing Meta into Stricter Digital Regulation
The European Commission has designated WhatsApp as a ‘very large online platform’ under the Digital Services Act after its Channels feature surpassed the 45 million‑user threshold in the EU. The move compels Meta to implement systemic‑risk assessments, stronger transparency and content‑mitigation measures within four months or face significant fines.
Microsoft’s Maia 200 Raises the Stakes in the Cloud AI Chip War
Microsoft has started deploying its Maia 200 AI accelerator built on TSMC 3nm, claiming substantial performance and cost advantages versus Amazon’s Trainium and Google’s TPU. The chip — designed to run large models efficiently at low power — is part of Microsoft’s strategy to secure more predictable, cheaper AI compute for Azure and to lessen reliance on Nvidia. An SDK preview is available to developers, while broader cloud rental availability is promised for the future.
China’s Lanqi Says It Has Built a PCIe 6.0 / CXL 3.0 Active Electrical Cable — A Step Toward Domestic High‑Speed Interconnects
Lanqi Technology says it has developed and system‑validated an active electrical cable compatible with PCIe 6.x and CXL 3.x, claiming a domestic first. The development addresses the signal‑conditioning challenges of next‑generation high‑speed links and signals growing Chinese capability in end‑to‑end datacentre interconnects, though production readiness and ecosystem interoperability remain open questions.
TikTok’s US Data‑Security Unit Blames Data‑Center Power Cut for Widespread System Glitches
TikTok’s US data‑security joint venture reported system failures after a power outage at a US data‑center partner site, restoring network connectivity but warning of lingering errors such as slow loads and incorrect engagement counts. The incident highlights operational vulnerabilities in data‑localization plans and shows how physical infrastructure dependencies—electricity and third‑party hosting—can undermine assurances about control and resilience.
Microsoft Unveils Maia 200 AI Chip to Wean Azure Off Nvidia
Microsoft has deployed Maia 200, its second‑generation AI chip built by TSMC, to some data centres and released developer control software, positioning the company to reduce dependence on Nvidia. Wider availability to Azure customers remains unspecified, but the move intensifies a trend of cloud providers building custom accelerators to control costs, supply risk and performance.