Hard Drives Are Making an AI Comeback. Yes, Hard Drives (www.wsj.com)

The AI boom is lifting an unexpected beneficiary: traditional hard drives. Western Digital and Seagate, the industry’s two largest HDD makers, each reported roughly 30% revenue growth in their latest quarters as demand from cloud and AI customers ramps up. While this isn’t an “Nvidia-style” breakout, investors are beginning to notice that large-scale AI workloads are driving substantial, sustained demand for high-capacity, low-cost storage. For the AI/ML community the resurgence matters because HDDs remain the most cost-effective medium for storing enormous training datasets, model checkpoints, and cold or nearline data that don’t require flash-speed access. As organizations expand data lakes, keep multiple model versions, and support massive inference repositories, high-capacity drives become a practical backbone—trading raw I/O latency for vastly lower $/TB. The result is a more tiered infrastructure where SSDs and GPUs handle hot, latency-sensitive work while HDDs absorb bulk storage needs, changing procurement, operational economics, and capacity planning for large-scale ML deployments.