**[Analysis: AI Data Growth Outpaces Compute Power Planning, Storage Becoming the New Bottleneck in AI Infrastructure]** BlockBeats News, August 15 — Western Digital's latest analysis indicates that as the scale of artificial intelligence applications rapidly expands, the construction of AI data centers is shifting from a sole focus on GPU compute power competition to a competition centered on data storage capabilities. Storage planning has now become a core aspect of AI infrastructure. The article cites IDC's forecast, which predicts that by 2030, the annual global data generation will reach 718ZB. Data generated by AI systems does not disappear after computational tasks are completed; training data, model checkpoints, embedding vectors, inference logs, prompts, output results, and evaluation data will continue to accumulate. Western Digital states that many current AI infrastructure plans overly emphasize GPU utilization while neglecting the data accumulation throughout the AI lifecycle. Data generated during training and inference processes will become critical assets for model iteration, quality evaluation, and compliance auditing. Storage costs will directly impact the long-term operational efficiency of AI systems. As data scales reach PB (petabyte) or even EB (exabyte) levels, a single storage architecture will struggle to meet demand. Enterprises need to adopt tiered storage strategies, utilizing high-performance flash storage for training and real-time inference, while employing high-capacity HDDs and object storage for long-term data preservation, historical records, and low-frequency access scenarios. The analysis suggests that in the future, the key metrics for AI infrastructure competition will not only be the number of GPUs but also the cost per PB of data storage, energy consumption, recovery efficiency, and data lifecycle management capabilities. If enterprises continue to view storage as a subsidiary aspect of computation, they may face issues such as uncontrolled data costs and reduced model iteration efficiency. [Original Link]