Google Unveils New AI Chips, Microsoft Opens Data Center Early, and NextDC's $2.2B Expansion - Silicon Brief #66
April 21, 2026 -
Silicon Brief
Hi there,
In this edition of Silicon Brief, we're diving into the latest developments in AI chips and data centers. Google is making waves with its collaboration with Marvell Technology to develop new AI chips aimed at enhancing model inference efficiency. This move is part of Google's broader strategy to expand its chip design partnerships. Meanwhile, Google is also set to launch a new generation of tensor processing units (TPUs) to compete with Nvidia in the AI chip market, following partnerships with Meta and Anthropic.
On the data center front, Microsoft has announced the early operational status of its Fairwater data center in Wisconsin, which is being hailed as the world's most powerful AI data center. Additionally, NextDC is planning to raise $2.2 billion to expand its data center capabilities to meet the growing demand from hyperscale and AI customers. However, a report by Bloomberg highlights that over half of planned data centers in the US are facing delays due to equipment shortages, impacting the expansion of AI infrastructure.
In this edition of Silicon Brief, we're diving into the latest developments in AI chips and data centers. Google is making waves with its collaboration with Marvell Technology to develop new AI chips aimed at enhancing model inference efficiency. This move is part of Google's broader strategy to expand its chip design partnerships. Meanwhile, Google is also set to launch a new generation of tensor processing units (TPUs) to compete with Nvidia in the AI chip market, following partnerships with Meta and Anthropic.
On the data center front, Microsoft has announced the early operational status of its Fairwater data center in Wisconsin, which is being hailed as the world's most powerful AI data center. Additionally, NextDC is planning to raise $2.2 billion to expand its data center capabilities to meet the growing demand from hyperscale and AI customers. However, a report by Bloomberg highlights that over half of planned data centers in the US are facing delays due to equipment shortages, impacting the expansion of AI infrastructure.
InPHRED Enters Data Center Optical Interconnect Market
InPHRED has announced its expansion into the data center optical interconnect market, focusing on high-speed connectivity solutions for AI infrastructure with its GaN micro-RC-LEDs and 1310 nm InP VCSELs. Read more
Steele Solutions Acquires Maysteel Industries
Steele Solutions has acquired Maysteel Industries to enhance its data center infrastructure capabilities. This move strengthens Steele Solutions' position in providing engineered products across various markets. Read more
Atlas Data Storage to Present at TSMC Symposium
Atlas Data Storage will showcase its CMOS-enabled molecular data storage technology at the TSMC Technology Symposium on April 22, 2026, highlighting its role in tackling enterprise IT challenges. Read more
Delta Unveils Solutions for AI Data Centers
At Data Center World 2026, Delta introduced its new infrastructure solutions for AI data centers, featuring integrated power, cooling, and intelligent controls to support AI workloads. Read more
Google Partners with Marvell for New AI Chips
Google is in discussions with Marvell Technology to develop two new AI chips to improve AI model inference efficiency. This initiative is part of Google's strategy to expand its chip design collaborations beyond Broadcom. Read more
SK hynix Mass Produces 192GB SOCAMM2 for AI Servers
SK hynix has initiated mass production of its 192GB SOCAMM2 memory, targeting improved performance for AI servers to address increasing AI application demands. Read more
RoboSense Unveils New Chipsets for 3D Perception
RoboSense has announced the EOCENE SPAD-SoC architecture and two new chipsets, Phoenix and Peacock, aimed at enhancing high-definition 3D perception. These innovations were revealed during the company's 2026 Tech Day in Shenzhen, China. Read more
Treon Launches AI Maintenance Solution on AWS
Treon has launched an AI-driven solution to optimize maintenance workflows on the AWS platform, aiming to improve efficiency and effectiveness. Read more
Google Unveils New AI Chips to Rival Nvidia
Google is set to launch a new generation of tensor processing units (TPUs) to boost AI inference capabilities, following collaborations with Meta and Anthropic. This initiative aims to enhance its competitiveness against Nvidia in the AI chip market. Read more
Sectra Secures £15.6 Million NHS Contract
Sectra has signed a £15.6 million contract to provide cloud imaging services to six NHS Trusts in London, focusing on radiology and breast imaging. Read more
US Data Centers Face Delays Due to Equipment Shortages
A report by Bloomberg reveals that over half of the planned data centers in the US are delayed due to shortages of critical electrical equipment, impacting AI infrastructure expansion. Read more
NextDC to Raise $2.2 Billion for Data Centre Expansion
NextDC is set to raise $2.2 billion to meet growing demand from hyperscale and AI customers. The company aims to fund this through an entitlement offer and hybrid securities expansion, following a significant increase in contracted utilisation. Read more
Microsoft's Fairwater Data Center in Wisconsin Goes Live Early
Microsoft CEO Satya Nadella has announced that the Fairwater data center in Mount Pleasant, Wisconsin, is operational ahead of schedule. The site, touted as the world's most powerful AI data center, has seen a significant investment increase to $7.3 billion. Read more
AlixLabs Secures €15 Million for Semiconductor Tech
AlixLabs, a Swedish semiconductor startup, has raised €15 million in Series A funding to advance its Atomic Layer Etching technology, enhancing efficiency and sustainability in semiconductor production. Read more
Anthropic Hires Former Google Executive for Energy Team
Sana Ouji has joined Anthropic to develop a global energy strategy for its data center expansion, after spending over six years at Google. Read more
NVIDIA to Showcase AI Innovations at COMPUTEX 2026
NVIDIA CEO Jensen Huang will present the company's latest AI advancements at COMPUTEX 2026. The keynote will highlight NVIDIA's 'Five-Layer Cake' approach and will be followed by GTC Taipei, featuring AI-related sessions and demos. Read more
Mountain Vector Energy Launches AI Solution for Cost Optimization
Mountain Vector Energy has launched Cufflink Nova™, an AI solution designed to help organizations in education, healthcare, and industrial sectors reduce utility costs and streamline operations. Read more
We hope you enjoyed this article.
Consider subscribing to one of our newsletters like Silicon Brief or Daily AI Brief.
Also, consider following us on social media:
More from: Data Centers
04/21
US Data Center Projects Face Widespread Delays Amid Equipment Shortages and Energy Concerns
04/21
Treon Introduces Make Cloud Solution for AI Maintenance on AWS
04/21
Six London NHS Trusts Adopt Sectra One Cloud for Shared Imaging Services
04/21
NVIDIA CEO Jensen Huang to Deliver Keynote at COMPUTEX 2026
04/21
RoboSense Unveils EOCENE SPAD SoC Architecture and Two Image Grade Chipsets
More from: Data Centers
04/21
US Data Center Projects Face Widespread Delays Amid Equipment Shortages and Energy Concerns
04/21
Former Google Energy Executive Joins Anthropic's New Energy Team
04/21
SK hynix Starts Mass Production of 192GB SOCAMM2 Memory for AI Servers
04/21
Google in Talks with Marvell to Develop New AI Inference Chips
04/20
Delta Electronics Unveils Integrated Power and Cooling Architecture for AI Data Centers
Subscribe to Silicon Brief
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.
