SanDisk's Platform & Emerging segment encompasses automotive storage, edge computing, and the nascent High Bandwidth Flash (HBF) product -- collectively generating ~$1.3B in FY2025 revenue. HBF is the most important emerging product: CEO Goeckeler described it as 'creating a new paradigm for AI inference solutions' that complements HBM (High Bandwidth Memory). If HBF gains adoption, it creates an entirely new addressable market that doesn't exist in current models. Automotive storage is growing steadily as vehicles require increasing data storage for ADAS, infotainment, and OTA updates. Edge computing storage serves IoT and on-premise AI inference deployments. This segment is small today but carries the highest optionality value.
CEO Goeckeler described HBF as 'creating a new paradigm for AI inference solutions' to complement HBM, positioning flash as a memory tier for AI inference workloads
High Bandwidth Flash (HBF) is a new product paradigm that positions flash memory as a complement to HBM for AI inference workloads. While HBM provides the highest bandwidth for GPU-adjacent memory, HBF aims to serve as a larger-capacity, lower-cost memory tier between HBM and standard SSDs. The use case is caching AI model parameters that don't fit entirely in HBM: for large language models with hundreds of billions of parameters, having a high-bandwidth flash tier could reduce inference latency and cost compared to reading from SSDs. If HBF gains adoption, it creates an entirely new TAM -- but this remains speculative. SK Hynix and Samsung are also exploring similar approaches, so SanDisk would not be alone if the market materializes.
HBF targets the gap between HBM bandwidth and SSD capacity, serving as a caching tier for AI inference model parameters that exceed HBM capacity