Who Feeds the GPUs? Inside AI's Hidden $30B Layer | Renen Hallak, VAST Data

Modern AI development demands a fundamental shift in infrastructure, moving beyond traditional data centers toward "AI factories" that prioritize massive data throughput and compute density. Renen Hallak, CEO of VAST Data, emphasizes that the crucial software infrastructure layer—the operating system for the AI era—must manage storage, networking, and compute to support the transition from training to complex agentic workflows. Enterprises require on-premise or private cloud environments to retain control over proprietary data and model weights, utilizing confidential computing to ensure security without sacrificing performance. As AI shifts toward autonomous agents capable of continuous fine-tuning and reinforcement learning, the ability to manage data, context windows, and model routing becomes the primary bottleneck. This evolution necessitates a unified, secure, and observable infrastructure capable of scaling to exabytes while maintaining strict access controls across distributed, multi-model environments.