Advancements in semiconductor technology, GPU development, and data center infrastructure that power AI computing capabilities.
Pittco Management has made an equity investment in UK-based Nscale Global Holdings, a company developing data centers for AI compute workloads, in partnership with Epping Forest Capital.
DEEPX and Sixfab have launched the DEEPX AI HAT+, an AI acceleration board for Raspberry Pi 5 that delivers low-power, high-performance inference capabilities for edge applications such as robotics and automation.
Qualcomm Technologies and Meta have entered into a long-term agreement for data center CPUs, with Qualcomm’s Dragonfly C1000 processor set to power Meta’s upcoming server fleet beginning in 2028.
Qualcomm has agreed to acquire AI software company Modular to strengthen its software foundation for generative and agentic AI across data center and edge environments. The deal is expected to close in the second half of 2026, subject to regulatory approval.
A new report from Exponential View finds that global AI revenue reached $25 billion in the first quarter of 2026, surpassing depreciation costs of major data center investments for the second consecutive quarter. The research suggests that demand for AI services is now large enough to sustain the industry's heavy infrastructure spending.
Advantech announced that its MIC-735 edge AI platform is undergoing validation through the NVIDIA Halos AI System Inspection Lab, making it one of the first platforms aligned with NVIDIA’s safety-focused framework for robotics and industrial systems.
Runpod has raised $100 million in a growth round led by Summit Partners, valuing the company at $1 billion. The funds will support expansion of its AI developer cloud platform used by over one million developers.
OpenAI and Broadcom have unveiled Jalapeño, a new inference processor designed specifically for large language models. The chip was developed in nine months and will be deployed at gigawatt scale in data centers starting in 2026.
SmartSens presented new automotive vision solutions at AutoSens USA 2026, including ADAS, surround-view, in-cabin, and driver monitoring systems, and previewed Flyingchip's upcoming M2 ISP chip.
kos.ai has launched general availability of its same day pay application review system powered by its virtual finance employee, Kos, following $12 million in inception financing.
SiMa.ai and Mistral Solutions have announced a partnership to develop an intelligent drone reference design combining SiMa.ai’s Physical AI platform with Mistral’s embedded systems expertise. The design aims to accelerate autonomous drone deployment and will be available to OEMs and system integrators in Q3 2026.
Odine has entered a strategic partnership with Supermicro to develop high-performance AI infrastructure in Türkiye, focusing on GPU systems, cloud architectures, and multi-cloud management solutions.
SPHERE AX and Blaize have signed a memorandum of understanding to jointly develop and commercialize AI semiconductor-based products, aiming to strengthen Korea-U.S. cooperation in the Physical AI sector.
Montana-Dakota Utilities has entered an electric service agreement with Applied Digital to supply power for the planned Polaris Forge 3 AI Factory near Center, North Dakota, which will require up to 430 megawatts of electricity when fully operational.
Sophia Space has chosen Apex to supply a satellite bus for a 2027 demonstration mission that will test its TILE orbital computing modules in real time.
At VivaTech 2026 in Paris, Envision announced Mission Gobi, a global initiative to develop 5 GW of low-carbon AI data centers in desert regions by 2030, integrating renewable energy, storage, and computing infrastructure.
Bell Canada, Cohere, Hypertec, and Buzz HPC have announced a large-scale collaboration to advance sovereign AI infrastructure in Canada, combining Bell’s data centers with Cohere’s AI models and Buzz HPC’s computing platform built on Hypertec hardware.
OpenMetal has expanded its v5 hardware catalog with two new private GPU server lines, the RP6000 and H200, built on NVIDIA RTX PRO 6000 Blackwell Server Edition and H200 NVL GPUs. The servers are designed for AI, machine learning, and high-performance computing workloads and are available now in the company’s US East data center.
Prometheus Hyperscale has received unanimous approval from Uinta County Commissioners for a conditional use permit to build its 1.25 gigawatt data center campus near Evanston, Wyoming. The facility will operate as an islanded microgrid powered by natural gas and is expected to begin construction within six months.
SK hynix has shipped samples of its new 12-layer HBM4E DRAM to major customers, featuring improved speed, power efficiency, and heat resistance for AI data centers and high-performance computing.
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.