Weekly Intelligence Brief on AI Infrastructure and Security Initiatives - Week of November 3, 2025

Key Takeaways

• Microchip Technology released Optical Ethernet PHY transceivers capable of 25 Gbps.
• NVIDIA partnered with SK Group to establish a large Artificial Intelligence (AI) factory in South Korea.
• NVIDIA also collaborates with Samsung on another AI factory to enhance semiconductor operations.
• Crusoe integrated NVIDIA BlueField-4 platform into Intelligent Transportation System (ITS) AI cloud for improved networking capabilities.
• Cisco and G42 work together to develop AI infrastructure in the UAE, addressing Graphics Processing Unit (GPU) (Graphics Processing Unit) capacity needs.

Microchip Technology introduced Optical Ethernet PHY transceivers, supporting speeds of up to 25 Gbps. This product addresses the demand for reliable networking solutions suitable for long distances, especially in industrial automation and telecommunications. The transceivers Converged Access Network (CAN) reach link lengths of 10 kilometers on Single-Mode Fiber (SMF) while integrating security features like IEEE 1588 PTP for synchronization.

NVIDIA Hardware Attestation Service (HAS) partnered with SK Group to create an AI factory in South Korea, featuring over 50,000 GPUs. The facility aims to enhance semiconductor research and enable advancements in digital twin technology and AI agent development. Meanwhile, in another partnership, NVIDIA and Samsung announced plans for a separate AI factory to optimize semiconductor manufacturing processes further.

Crusoe upgraded ITS AI cloud infrastructure by integrating the NVIDIA BlueField-4 platform, enhancing multitenancy and securing high-performance networking. The upgrade focuses on providing zero-trust isolation among virtual instances. Additionally, Cisco expanded ITS partnership with G42 to develop AI infrastructure in the UAE, integrating advanced technologies to support the region's initiatives for digital transformation.

1. Microchip Technology introduces Optical Ethernet PHY transceivers supporting up to 25 Gbps with PTP and Media Access Control Security (MACsec).
   Microchip Technology introduced Optical Ethernet PHY transceivers up to 25 Gbps, addressing the demand for reliable connectivity.
2. NVIDIA and SK Group establish AI factory in South Korea
   NVIDIA partners with SK Group to establish an AI factory in South Korea with 50,000 GPUs to enhance semiconductor research.
3. NVIDIA and Samsung to Construct AI Factory in South Korea
   NVIDIA and Samsung announced a new AI factory using 50,000 NVIDIA GPUs to enhance AI-driven semiconductor manufacturing.
4. Itential Model Context Protocol (MCP) secures AI interactions and infrastructure
   Itential MCP provides a structured framework that safeguards infrastructure from uncontrolled AI interactions, ensuring secure automation.
5. Agentic AI and the Future of Infrastructure
   AI integration is prompting infrastructure leaders to rethink traditional operational frameworks.
6. H.R.H. Prince Abdulaziz bin Salman attends Cisco's and King Abdullah University of Science and Technology's (KAUST) launch of landmark AI Institute to accelerate AI research, development and talent in Saudi Arabia.
   Cisco and KAUST launched a new AI Institute to advance AI research and workforce skills, supporting Saudi Arabia's Vision 2030 goals.
7. ASUS Launches XA NB3I-E12 AI Server Built with NVIDIA HGX B300
   ASUS launched the XA NB3I-E12 AI server with eight NVIDIA GPUs for enterprise AI workloads, enhancing performance and scalability.
8. Networked Control System (NCS) Receives Frost & Sullivan's 2025 Singapore Customer Value Leadership Recognition in the Contact Center Technology Services Industry
   NCS received Frost & Sullivan's 2025 Customer Value Leadership recognition for AI advancements in contact centers.
9. Crusoe integrates NVIDIA BlueField-4 for enhanced AI capabilities
   Crusoe integrates NVIDIA BlueField-4 into ITS AI cloud platform, enhancing multitenancy and enabling high-performance 800Gb/s networking.
10. Xage Security integrates NVIDIA BlueField to deliver unified Zero Trust for AI
    The integration aims to deliver low-latency security for AI environments, ensuring compliance with key regulations.
11. MITRE develops AI-native network architecture for telecom
    MITRE aims to ensure that all intellectual property developed in the AI-native initiative will be available to trusted U.S. industry.
12. Check Point partners with NVIDIA for AI Cloud Protect security solution
    AI Cloud Protect secures AI infrastructure against cyber threats, with Check Point and NVIDIA offering seamless performance without impact.
13. Akamai Technologies launches AI Inference Cloud at the edge
    Akamai Technologies launched Akamai Inference Cloud, enabling global AI inference with low latency from the edge of the internet.
14. Cisco and G42 deepen US-UAE technology partnership to build secure, end-to-end AI infrastructure in the UAE.
    Cisco and G42 aim to address GPU capacity challenges in the UAE, with plans for large-scale AI deployments using advanced technologies.
15. Organizations Gain Advantage by Adopting Automation in 2025
    Leaders Agent Runtime Environment (ARE) encouraged to integrate AI into orchestration, optimizing decision-making processes.
16. Automated Model Deployment (AMD) and DOE partner for AI supercomputers lux and discovery
    AMD and the U.S. Department of Energy announced the Lux and Discovery supercomputers to boost U.S. AI and High performance computing (HPC) leadership.