Networking for AI: Building the foundation for real-time intelligence

Networking for AI: Building the foundation for real-time intelligence

To address the complexities of IT management, the Ryder Cup collaborated with technology partner HPE to establish a central operational hub. This hub features a platform that allows tournament staff to access data visualizations, aiding in operational decision-making. The dashboard utilizes a high-performance, private-cloud environment to compile insights from various real-time data feeds.

Jon Green, CTO of HPE Networking, emphasizes the necessity of effective networking for successful artificial intelligence (AI) implementations. He notes that a “disconnected AI” system lacks effectiveness, stressing the importance of a robust network for data movement essential for both training and inference.

As organizations increasingly adopt distributed, real-time AI applications, future networks must be capable of handling exponentially larger volumes of data at quicker speeds. The experience at Bethpage Black highlights a broader lesson: networks optimized for inference are critical to realizing AI’s potential in practical applications.

Despite advancements, more than half of organizations face challenges in operationalizing their data pipelines. A recent HPE survey revealed that 45% of 1,775 IT leaders indicated they could manage real-time data for innovation, a remarkable increase from just 7% the previous year. However, gaps remain in connecting data collection with real-time decision-making processes.

The significance of network infrastructure design is highlighted as a potential solution. Traditional enterprise networks often cater to predictable application flows but are typically not equipped for the dynamic data movement required for AI, especially in tasks that demand the rapid transfer of large datasets among GPUs.

Green pointed out the distinct performance needs of AI networks, such as ultra-low latency, lossless throughput, and specialized equipment. During the Ryder Cup, a Connected Intelligence Center was established to collect and analyze data, including ticket scans, weather conditions, and spectator movement. This center utilized 67 AI-enabled cameras and provided staff with an immediate overview of event operations through an operational intelligence dashboard.

Source: https://www.technologyreview.com/2025/11/18/1127997/networking-for-ai-building-the-foundation-for-real-time-intelligence/

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top