Western Digital CPO: Flash for Now, HDD for Long-Term AI Storage

ChainCatcher report: Ahmed Shihab, Chief Product Officer at Western Digital, posted that the key to AI storage competition is not merely pursuing the fastest media, but rather the ability to scale capacity sustainably at an affordable cost. He noted that many architectures function well initially, but when data scales from several petabytes to hundreds of petabytes or even exabytes, cost overruns may expose critical flaws. Flash storage is ideal for high-performance, low-latency use cases such as model weights, GPU overflow, KV caches, and session contexts; HDDs are better suited for large-scale, long-term, and cost-sensitive data such as training corpora, logs, checkpoints, compliance records, synthetic data, and inference outputs. He concluded: “Flash handles the present; HDD handles the entire lifecycle.” Under AI-scale demands, storage cost itself becomes an architectural issue. Persistently storing massive volumes of data on high-performance media will crowd out budgets for compute, networking, power, and personnel. For many bulk storage workloads, the question is not whether flash can store the data—but whether customers can afford the long-term cost of using flash exclusively. The future of AI storage will not be dominated by a single technology; instead, it must be designed in layers based on performance, cost, power consumption, density, reliability, and data lifecycle. He emphasized that this is not a competition between flash and HDD, but rather a need to match the right medium to each workload from the outset—otherwise, architectures may compromise business sustainability as they scale.