IEEE Life Fellow Tom Coughlin on why memory and storage capacity will shape AI data center growth

Tom Coughlin, an IEEE Life Fellow and president of Coughlin Associates, emphasizes that memory and storage capacity are critical factors shaping the future growth of AI data centers. With over four decades in digital storage, Coughlin highlights that storage is not just a passive repository but the foundational infrastructure enabling computation, AI, and digital society. He views the current AI boom as a potential inflection point in computing history, particularly if the focus shifts toward augmented human intelligence, which he believes can empower people to achieve more and improve quality of life. Coughlin warns that despite the excitement around AI compute power, there is a growing risk of memory and storage shortages that could constrain data center expansion and other applications from late 2025 through 2026. The surge in demand stems from the need to store vast amounts of raw sensor data, AI training models, logs, and domain-specific data to improve inference results. Emerging technologies like Inference Context Aware Storage are expected to further increase storage requirements.