
Nvidia Eyes Lepton AI, Apple Invests $1B in AI Servers, Alibaba Warns of Overheating Market - Silicon Brief #11
Hi there,
In this edition of Silicon Brief, we're diving into the latest developments in AI chips and data centers. Nvidia is reportedly in talks to acquire Lepton AI, a move that could further bolster its position in the AI hardware market. Meanwhile, Apple has placed a significant order of $1 billion for Nvidia servers to enhance its AI capabilities. On the data center front, Alibaba has issued a warning about the potential overheating of the AI infrastructure market, highlighting the rapid pace of development in this sector.
Additionally, Microsoft and Core42 are partnering with Abu Dhabi to develop AI infrastructure, showcasing the global expansion of AI data centers. In Japan, alt Inc. and Highreso Co. Ltd. have launched a new GPU cloud, further expanding the country's AI capabilities. These stories underscore the dynamic and rapidly evolving landscape of AI technology and infrastructure.
Energy & Sustainability
Lumentum joins NVIDIA's AI Photonics Ecosystem. GMI Cloud unveils AI Inference Engine for businesses. Taiwan's AI supply chain showcased at NVIDIA GTC. alt Inc. and Highreso Co. Ltd. launch GPU Cloud in Japan. Mitsubishi and JFE to build AI data center in Kawasaki. Switch, Ltd. completes $3.5 billion securitized debt financings. Alibaba exec warns of AI infrastructure market overheating. Segro and Pure DC to build 56MW data center in London. Kasm Technologies offers AI workspaces on Oracle Cloud. Treasure Global Inc. launches AI Cloud in Malaysia. Cinnova Technologies partners with Microsoft and AWS. Algeria begins construction of AI data center in Oran. Microsoft and Core42 partner with Abu Dhabi for AI infrastructure. OpenAI plans dedicated data center for storage. Scan Computers chooses Peak:AIO for GPUaaS expansion. Oracle to develop 30,000 AMD GPU cluster.
- Oklo Increases Aurora Reactor Capacity for AI Data Centers
- Schneider Electric to Invest $700 Million in U.S. for Energy and AI
Technology
OpenLight has announced the availability of samples for its 1.6Tb DR8 Photonic Integrated Circuit, aimed at datacenter networking and AI applications. Magic Software Japan and Visnu have formed a strategic partnership to integrate their technologies for enhanced real-time data analysis. At Nvidia's GTC 2025 conference, Solidigm introduced the D7-PS1010, a liquid-cooled SSD for AI data centers. CIQ has launched federation capabilities in its Fuzzball platform for managing AI workloads. Corning Incorporated unveiled GlassWorks AI™, enhancing AI data center infrastructure.
- OpenLight Releases 1.6Tb DR8 Photonic Integrated Circuit Samples
- Solidigm Introduces Liquid-Cooled SSD for AI Data Centers
- CIQ Introduces Fuzzball Federation for AI Workloads
- Corning Launches GlassWorks AI for Data Centers
- NEXCOM Launches NSA 5200 for AI Cybersecurity
- Magic Software Japan and Visnu Partner for Edge AI Integration
Regulation & Policy
Cerebras Systems experiences delays in its IPO due to a national security review linked to a significant investment from Abu Dhabi-based G42.
Data Centers
Lumentum Holdings Inc. has joined NVIDIA’s AI Photonics Ecosystem to improve data center speed and efficiency. GMI Cloud has launched a new AI Inference Engine for businesses, enhancing AI application deployment. Taiwan showcased its AI supply chain at the NVIDIA GTC, highlighting the Asia New Bay Area as a global innovation hub.
- Lumentum Joins NVIDIA's AI Photonics Ecosystem
- GMI Cloud Introduces AI Inference Engine for Businesses
- Taiwan's AI Supply Chain Debuts at NVIDIA GTC
- alt Inc. and Highreso Co. Ltd. Launch GPU Cloud in Japan
- Mitsubishi and JFE to Build AI Data Center in Kawasaki
- Switch, Ltd. Completes $3.5 Billion Securitized Debt Financings
- Alibaba Exec Warns of AI Infrastructure Market Overheating
- Segro and Pure DC to Build 56MW Data Center in London
- Kasm Technologies Launches AI Workspaces on Oracle Cloud
- Treasure Global Inc. Launches AI Cloud in Malaysia
- Cinnova Partners with Microsoft and AWS
- Algeria Begins Construction of AI Data Center in Oran
- Microsoft and Core42 to Develop AI Infrastructure for Abu Dhabi
- OpenAI Plans Dedicated Data Center for Storage
- Scan Computers Expands GPUaaS with Peak:AIO Storage
- Oracle to Develop 30,000 AMD GPU Cluster
- Cassava Technologies to Launch AI Factory in Africa
- Data Cities to Build AI Data Centers in California
- Eurazeo and Metrobloks Partner for €100M Data Center Development
- Crusoe Energy Expands Stargate Data Center in Texas
Chips & Hardware
NATO Innovation Fund has co-led a €25 million Series A funding round for Camgraphic to boost R&D and establish a pilot manufacturing line in Italy. On March 27, 2025, Akamai Technologies launched its Cloud Accelerated Compute Instances using NETINT VPUs to enhance video streaming efficiency. MemryX secured $44 million in Series B funding to advance its Edge AI semiconductor solutions, focusing on the MX3 Accelerator chip and MX4 development.
- NATO Innovation Fund Invests in Camgraphic
- Akamai Launches Cloud Compute Instances with NETINT VPUs
- MemryX Secures $44 Million for Edge AI Solutions
- Nvidia in Talks to Acquire Lepton AI
- Rapidus Partners with Quest Global for 2nm AI Semiconductors
- IDTechEx Predicts Significant Growth in HBM Demand by 2035
- H3C Warns of Nvidia H20 Chip Shortage in China
- Apple Orders $1 Billion in Nvidia Servers for AI
- NUS Develops New Transistor for AI
Subscribe to Silicon Brief
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.