We’re pouring more and morepower into copper just to move bits a few meters – and paying for it in cooling. Silicon photonics slashes interconnect power per bit, easing the power/cooling crunch and freeing budget for compute, not heat. #Datacenters# #EnergyEfficiency#

Getting to 1000x energy efficiency in AI isn’t about one breakthrough. It’s about solving two hard constraints: 1. Data movement dominates energy 2. Amdahl’s Law caps system-level gains Which means you have to rethink everything: models, hardware, and how they’re designed together. If this kind of problem excites you, you’ll enjoy our latest blog:

TPU 8 will be a massive leap forward in energy efficiency—with up to 2x better PPW than Ironwood. But for us, sustainability isn’t about waiting for the next best chip; it’s a daily effort to make our entire production fleet more efficient → #8daysofTPU8#

Every additional watt consumed inside a data center increases cooling requirements. That's why #OpticalInterconnects# are becoming essential for scaling AI infrastructure while improving energy efficiency. Discover how our DFB #LaserArrays# are powering the AI Infrastructure of tomorrow!

CHINA isn’t just making display screens — it’s building astonishing technologies that are waterproof, shatter‑resistant, stretchable and compressible, with ultra‑vivid colors and exceptional energy efficiency… and the features are far too many to count.

The next era of #AI# #infrastructure# will be defined by more than scale. It requires resilience, energy efficiency, and intelligent data center architecture built for sustained high-performance workloads. At Oslo Tech Show 2026, Haakon Bryhni, Head of Tydal Data Center under Bitdeer Technologies Group, will share his perspective on resilient networks, edge computing, and the evolution of next-generation data centers supporting AI at scale. With a background spanning research and real-world deployment, his insights highlight how infrastructure design is becoming a critical factor in AI performance and reliability. A valuable session for those building, operating, or scaling AI infrastructure. Don’t miss it ✨: #AIInfrastructure# #DataCenter# #neocloud# #bitdeerai#

PyTorch Foundation supported the ExecuTorch Hackathon in San Francisco, where more than 100 participants across 20+ teams built real-time AI applications using PyTorch and ExecuTorch. Teams built on Snapdragon-powered Samsung Electronics Galaxy S25 Ultra devices, focusing on latency, offline capability, privacy-sensitive processing, energy efficiency, and real-time user experience. Congratulations to the winning teams: 1st Place: SafeScreen AI, an on-device visual safety layer 2nd Place: SixthSense, an assistive wearable that converts visual information into directional haptic signals 3rd Place: Toddle AI, a privacy-first prototype for analyzing toddler walking patterns locally The winning projects showed how local execution can support applications that require immediate feedback, limited connectivity, or sensitive data processing. Read the full recap from @matthew_d_white (PyTorch Foundation), Andrew Caples (@Meta), and Lauren Lunde (@Qualcomm):

China has launched the world’s first commercial underwater data center powered by offshore wind near Shanghai’s Lingang area. This 24-megawatt facility is now fully operational with nearly 2,000 servers for AI tasks, big data, 5G networks, self-driving cars, and smart robots. The sealed units sit about 10 meters underwater and 10 kilometers offshore next to a large wind farm. More than 95% of the power comes directly from the wind turbines via special cables that reduce power loss, while seawater cools the servers naturally through heat exchangers. This eliminates the need for large cooling systems or fresh water, giving it an energy efficiency rating of about 1.15 compared to 1.5 to 2.0 for typical land data centers. The project progressed rapidly: the deal was signed in June 2025, construction ended by October 2025, tests ran in February 2026, and full service started in mid-May 2026, at a total cost of about 1.6 billion yuan. This solves two major data center problems : high electricity use from AI and heavy water consumption for cooling by combining clean wind power with the sea’s natural cooling.