Nvidia invests $2B as AMD partners with IBM Cloud

Key Takeaways

• Nvidia is investing $2 billion in photonics technology to address AI infrastructure bottlenecks and speed up data transfer.
• IBM Cloud and AMD are partnering to provide integrated hardware and software solutions for complex AI and high-performance computing workloads.
• Oracle views communication networks as essential AI infrastructure, critical for integrating AI across systems and managing demands from gigawatt-scale data centers.
• SK Telecom, Supermicro, and Schneider Electric are collaborating on modular AI data centers for faster, more scalable deployment.
• Huawei introduced new AI Data Center products, including the AI Data Platform and Xinghe AI Fabric 2.0, at MWC 2026 to accelerate AI adoption.
• Circana launched its Complete Why AI solution to provide CPG companies with detailed sales performance insights by analyzing up to 60 drivers.
• AI has the potential to improve digital product accessibility for 15-20% of users with cognitive or sensory differences through adaptive experiences.
• New York City schools are developing plans to integrate AI into K-12 education, focusing on student preparation and teacher training.
• Agentic AI, capable of autonomous planning and action, is gaining traction, with a new course starting March 9, 2026, distinguishing it from tools like ChatGPT.
• Chinese AI startups are increasingly focusing on international markets, forming global partnerships and expanding their reach worldwide.

Huawei unveils AI Data Centers and new solutions at MWC 2026

At MWC Barcelona 2026, Huawei introduced new AI Data Center products and solutions, including the Huawei AI Data Platform and Xinghe AI Fabric 2.0. These innovations aim to speed up businesses' adoption of AI technology. The AI Data Platform helps enterprise AI agents use knowledge more effectively and speed up responses. Xinghe AI Fabric 2.0 is an upgraded solution designed to maximize computing power and ensure continuous services for businesses.

Cloud AI computing needs innovative hardware and software

The growing complexity of AI and high-performance computing workloads requires advanced infrastructure. Organizations are increasing their investment in AI and HPC, with many planning significant expansions. Hybrid cloud and on-premises environments are common, but managing this complexity requires new software tools and expertise. Strategic partnerships, like IBM Cloud with AMD, aim to provide integrated solutions that balance performance, cost, and flexibility for AI workloads.

Cloud AI computing needs innovative hardware and software

The increasing complexity of AI and high-performance computing (HPC) workloads is driving demand for new infrastructure solutions. A recent study shows that most organizations are expanding their AI infrastructure, with many planning to integrate HPC workloads. Hybrid cloud and on-premises environments are becoming standard, but this complexity requires advanced software tools for management. Partnerships like IBM Cloud and AMD are developing integrated solutions to meet the evolving needs for performance, cost efficiency, and flexibility in AI computing.

Oracle sees communications as key AI infrastructure

Oracle is now viewing communication networks as a fundamental part of AI infrastructure, similar to energy and construction. The company believes that integrating AI across networks, customer systems, and back-office platforms is crucial. The rise of massive gigawatt-scale data centers is creating new demands on networks, utilities, and billing systems. Oracle emphasizes that communications providers have unique visibility into global data flows, which can be leveraged for AI-driven insights if data fragmentation issues are addressed.

Oracle sees communications as key AI infrastructure

At MWC 2026, Oracle is redefining communications networks as essential AI infrastructure, not just a separate industry. The company highlights the critical need for AI-driven data integration across networks, customer systems, and back-office platforms. The growth of gigawatt-scale data centers presents new challenges for networks, utilities, and billing. Oracle points out that communication networks offer unparalleled visibility into global data, which can be used for AI insights if data silos are broken down.

Nvidia invests $2 billion in photonics for AI infrastructure

Nvidia is investing $2 billion in photonics technology to address bottlenecks in AI infrastructure. This investment, along with supply agreements for optical components, will support Lumentum Holdings. Photonics uses light to transmit data faster and more efficiently, which is crucial for large AI computing systems. The collaboration will focus on silicon photonics to speed up data transfer within AI systems. This move aims to boost manufacturing capacity and accelerate the development of next-generation optical components for AI.

Circana launches AI solution for CPG sales insights

Circana has launched its Complete Why solution, using AI to help consumer packaged goods (CPG) companies understand sales performance. This tool analyzes up to 60 drivers per product, including price, promotions, competition, and weather, at the store and week level. The solution is integrated into Circana's Unify+ platform for easy visualization and fast insights. It aims to provide CPG companies with accurate data attribution to make smarter, quicker decisions in a challenging market.

New York schools consider AI's future role

New York City's school system is deciding how to incorporate artificial intelligence into education. While many school systems have announced large-scale AI adoptions, New York has been slower to commit. The city aims to prepare students for AI-powered lives and careers and train teachers on responsible AI use. The school system is expected to announce its plans soon regarding the role of AI in K-12 education.

AI could help 1 in 5 users access products

Many digital products are designed for a mythical 'normal user,' leaving about 15-20 percent of the population with cognitive, attentional, or sensory processing differences struggling to use them. Traditional accessibility efforts have focused on physical or sensory needs, overlooking cognitive variability. AI offers a potential solution by enabling adaptive experiences that cater to a wider range of users. This technology could finally help create truly universal digital products that accommodate diverse human needs.

New RFP template guides AI usage control and governance

Security leaders are now getting budgets to secure AI, but many organizations lack clear requirements for AI Governance. A new RFP template helps evaluate AI Usage Control (AUC) solutions, focusing on 'interaction-level inspection' rather than just cataloging apps. This approach allows control over AI interactions regardless of the specific tool used, preventing security from becoming a bottleneck. The template guides vendors to prove their ability to detect AI usage in various scenarios and outlines eight key domains for mature AI governance.

Chinese AI startups focus on global markets

Many Chinese AI startups are prioritizing international markets, with companies like Tripo AI reporting 90% of their users are outside China. These startups are forming strategic partnerships with global corporations and developing tools to help Chinese companies sell overseas. Businesses in Europe and the U.S. are more willing to adopt new AI tools, even without immediate revenue gains. Chinese AI companies are rapidly expanding worldwide, with some aiming to compete with major global players.

SK Telecom plans modular AI data centers with partners

SK Telecom, Supermicro, and Schneider Electric are collaborating to create modular AI data centers. This partnership aims to speed up AI data center construction and reduce supply chain issues. They will develop pre-fabricated modules that combine AI servers with power and cooling systems, allowing for building-block construction. This approach offers faster deployment, cost efficiency, and scalability compared to traditional methods. SK Telecom will provide operational expertise, Supermicro will supply GPU servers, and Schneider Electric will handle MEP infrastructure.

Agentic AI course starts March 9, 2026

A new course on Agentic AI, a type of artificial intelligence that can plan and act autonomously, begins on March 9, 2026. Unlike tools like ChatGPT, Agentic AI systems can break down complex goals into tasks and use external tools to achieve them. These systems improve over time through memory, reasoning, and feedback loops. Agentic AI is being used in various industries like finance, retail, and healthcare to automate workflows and create intelligent assistants.

Sources

• Huawei Releases AI DC Products and Solutions for Smart Future of Shared Success
• Enabling Advanced AI Computing in the Cloud with Innovative Hardware/Software Collaborations
• Enabling Advanced AI Computing in the Cloud with Innovative Hardware/Software Collaborations
• MWC 2026: Oracle positions communications as core AI infrastructure
• MWC 2026: Oracle positions communications as core AI infrastructure
• Nvidia Targets AI Infrastructure Bottleneck With $2 Billion Photonics Investment
• Circana Launches Complete Why Solution with AI-Enabled Insights to Decode CPG Sales Performance
• A.I. in New York Schools: What Lies Ahead?
• 1 in 5 Users Can’t Fully Use Your Product. AI Might Finally Fix That.
• New RFP Template for AI Usage Control and AI Governance
• CNBC's China Connection newsletter: New AI players think global from day one
• SK Telecom and partners plan modular AI data centers built like building blocks
• AGENTIC AI Course Starting on Mon, Mar 9, 2026