Advancements in semiconductor technology, GPU development, and data center infrastructure that power AI computing capabilities.
The Open Compute Project Foundation is redesigning data center power architecture with 1 Megawatt racks to reduce energy losses from 40% to 7%, according to a press release by DIGITIMES ASIA.
Qualcomm is re-entering the data center CPU market with custom processors that integrate Nvidia technology for enhanced AI capabilities.
Fabric8Labs and Wiwynn will demonstrate advanced ECAM-enabled cold plates at Computex 2025, enhancing cooling for AI data centers.
Vertiv has announced its strategic alignment with NVIDIA's AI roadmap to deploy 800 VDC power architectures for next-generation data centers, with solutions planned for 2026.
Supermicro is now accepting orders for over 20 systems optimized for NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, enhancing enterprise AI capabilities.
Marvell Technology and NVIDIA have announced a collaboration to offer custom solutions for advanced AI infrastructure, utilizing NVIDIA's NVLink Fusion technology.
Wiwynn has showcased its latest AI servers and cooling solutions at Computex 2025, featuring NVIDIA and AMD technologies.
NVIDIA has unveiled NVLink Fusion, a new silicon solution that allows industries to build semi-custom AI infrastructure with a wide range of partners. This initiative enables integration of custom CPUs and AI accelerators with NVIDIA's products.
Axiado and Posiflex Group have introduced an AI-enabled Baseboard Management Controller solution for Intel servers at COMPUTEX 2025, enhancing security and efficiency.
OpenAI is set to be a primary tenant in a 5-gigawatt data center campus in Abu Dhabi, developed with G42, potentially becoming one of the world's largest AI infrastructure projects.
G42 and iGenius have announced a partnership to develop a major AI supercomputer in Italy, with a $1 billion investment over five years.
Cerebras Systems has introduced the Qwen3-32B model on its inference platform, offering real-time AI reasoning at unprecedented speeds.
CIQ has launched Rocky Linux from CIQ for AI, a Linux distribution optimized for artificial intelligence applications, offering enhanced performance and security.
Ampere has launched the Ampere Systems Builders program and expanded its AmpereOne M platforms, enhancing AI and cloud infrastructure.
Achronix Semiconductor Corporation has launched the VectorPath 815, a new FPGA accelerator card designed to enhance AI inferencing and high-performance computing. The card is powered by the Speedster 7t1500 FPGA, offering significant performance improvements over traditional GPU solutions.
Supermicro has announced its Data Center Building Block Solutions (DCBBS), designed to simplify and expedite the construction of AI/IT liquid-cooled data centers.
Delta Electronics has introduced its latest AI data center networking solutions, including a 51.2T Co-Packaged Optics Ethernet switch, at COMPUTEX 2025.
atNorth will host the infrastructure for Sweden's National AI Cloud at its SWE01 data center in Stockholm, in partnership with 6G AI Sweden, using NVIDIA's latest AI chips.
TrendForce reports that US and Chinese cloud service providers are rapidly developing in-house AI chips to reduce reliance on imports amid geopolitical tensions.
NeuReality has announced that its NR1 AI Inference Appliance now comes preloaded with popular AI models, offering improved deployment speed and cost efficiency.
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.
