The 512TB E2 R6060 stands as the centerpiece of DapuStor’s new lineup, effectively doubling the capacity of its previous 245TB iteration. Designed specifically for scale-out storage environments, the drive aims to address the massive data requirements of modern AI training. Beyond raw capacity, the company introduced specialized form factors including the R6 E1.S liquid-cooled SSD, which integrates cold-plate technology to maintain thermal efficiency in high-density GPU servers running GPUDirect Storage workloads.

Industry recognition followed the product showcase, as a collaborative project between DapuStor and UBIX earned finalist status in the FMS 2026 Best of Show Device Engineering category. This solution, utilizing R6100 PCIe 5.0 TLC SSDs, previously demonstrated high-performance metrics in MLPerf Storage v2.0 testing. During the event, DapuStor VP of R&D Xiang Chen and Xiaohongshu technical lead Zhenyu Shu addressed the technical hurdles of AI infrastructure, focusing on Flexible Data Placement, cache management, and power-loss protection in large-capacity QLC drives.