Share
Linkedin
Twitter
Facebook
Whatsapp
Copy Link
Samsung is close to securing Nvidia certification for its HBM4 AI memory chips, entering final qualification phases after supplying initial samples in September. The company prepares for mass production in February, aiming to debut alongside Nvidia's Rubin AI platform at GTC 2026. Meanwhile, SK Hynix reports significant progress in its own HBM4 tests, setting up an intense competition between South Korean rivals.
Google DeepMind CEO Demis Hassabis told the Financial Times that while the artificial intelligence industry isn't in a bubble, certain pockets show clear signs of excess. He specifically called out multi-billion-dollar seed rounds for early-stage AI companies without proven products as unsustainable, suggesting market corrections may be ahead.
AMD is reportedly developing the Ryzen AI MAX 400 Series, codenamed 'Gorgon Halo,' as a refresh of its Strix Halo processors. The new lineup features boosted CPU and GPU clock speeds, support for faster LPDDR5X-8533 memory, and the same Zen 5 architecture. The flagship Ryzen AI MAX+ 495 leads with 16 cores and clock speeds up to 5.2 GHz, targeting AI workloads and integrated graphics performance.
Intel is reallocating manufacturing resources from consumer processors to data center chips as unexpected AI infrastructure demand creates supply bottlenecks. The company admits hyperscaler customers caught it off guard with surging orders, forcing difficult choices that could impact PC availability and pricing through early 2026.
Austin-based Neurophos has raised $110 million in Series A funding led by Bill Gates' Gates Frontier to develop optical processing units that promise 100x better performance than current AI chips. The startup's breakthrough uses metamaterial optical transistors 10,000 times smaller than existing photonic elements, aiming to address AI's mounting energy and compute challenges.
Nvidia is reportedly cutting production of its RTX 50-series consumer graphics cards by 15-20% through at least Q3 2026 to meet surging AI demand. The RTX 5060 may be discontinued for six months, while flagship models like the RTX 5090 and 5070 Ti could become nearly unobtainable. Industry sources warn of potential 30-50% price increases as the company prioritizes AI data centers over gaming hardware.
Alibaba is preparing to list its AI chipmaking unit T-Head through an initial public offering, tapping surging investor demand for companies competing in the AI accelerator market. The move follows CEO Eddie Wu's pledge of over $53 billion toward AI infrastructure investments. T-Head will first be restructured with partial employee ownership before the IPO, though timing remains unclear.
OpenAI CEO Sam Altman is courting sovereign wealth funds in the Middle East for a massive $50 billion funding round that could value the company at up to $830 billion. The ChatGPT maker, still unprofitable, has committed over $1.4 trillion to AI infrastructure spending while facing intensifying competition from Anthropic and Google in the artificial intelligence race.
AMD released Adrenalin 26.1.1 with an optional AI Bundle featuring PyTorch, ComfyUI, Ollama, LM Studio, and Amuse for streamlined local AI workflows. The WHQL driver adds support for Ryzen AI 400 series processors and delivers game optimizations for Starsand Island and Avatar: Frontiers of Pandora, alongside fixes for graphics and performance bugs across Radeon GPUs.
Broadcom is cementing its position as the go-to partner for custom AI chips, with analysts projecting 60% market share in AI server compute ASIC design through 2027. CEO Hock Tan reports insatiable demand across both chip and networking divisions, with Google alone planning to deploy 6-7 million TPU processors in 2027. The company holds an 18-month lead over customer-owned chip programs, positioning it to capture explosive AI infrastructure spending growth.
AI networking startup Upscale AI raised $200 million in Series A funding led by Tiger Global, Premji Invest, and Xora Innovation, pushing its valuation past $1 billion. The company is developing SkyHammer chips designed specifically for scale-up AI networking to compete with Nvidia's NVSwitch and tackle network congestion plaguing AI data centers.
Volvo unveiled its EX60, the world's first electric SUV powered by Google Gemini AI. Built on the new SPA3 platform with megacasting technology, the midsize electric SUV offers up to 400 miles of range and 670 horsepower. The EX60 integrates NVIDIA and Qualcomm chipsets for AI-native computing and arrives at U.S. dealers starting summer 2026.
Nvidia has significantly upgraded its Vera Rubin AI accelerator specifications in response to competitive pressure from AMD's Instinct MI455X. The chip now features 22.2TB/s memory bandwidth—a 10% increase—and a 2.3kW TDP, up from the originally announced 1.8kW. These Vera Rubin upgrades aim to secure hyperscaler adoption and maintain Nvidia's dominance in AI datacenter infrastructure.
Nvidia CEO Jensen Huang argued at the World Economic Forum that AI will drive demand for plumbers, electricians, and construction workers as the industry undertakes what he calls the largest infrastructure buildout in human history. While AI threatens white-collar displacement, Huang says trade workers could earn six-figure salaries building chip factories and data centers needed to power the AI boom.
The House Foreign Affairs Committee voted 42-2 to advance the AI Overwatch Act, which would give Congress veto power over AI chip exports to China and other adversaries. The bipartisan legislation challenges Trump's decision to allow Nvidia H200 sales to China with a 25% revenue cut, creating a rift between China hawks in Congress and the White House over national security versus technological dominance.
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.