Sandisk unveils 256 TB UltraQLC NVMe SSD for enterprise workloads

Sandisk unveils its 256 TB UltraQLC NVMe SSD, pushing boundaries for Artificial Intelligence and data-driven enterprises with record capacity and efficiency.

Sandisk has demonstrated its latest technological milestone: a high-capacity 256 TB NVMe enterprise solid-state drive, powered by the new UltraQLC technology platform. This SSD represents a significant leap in both storage scale and architecture, paving the way for next-generation enterprise solutions. Built on a foundation combining BiCS8 QLC CBA NAND, proprietary controllers and intricate system enhancements, the UltraQLC platform enables unprecedented capacity, performance, and energy efficiency for demanding data applications.

With the accelerating demands of the Artificial Intelligence era, enterprises face increasingly complex and customizable storage needs. Sandisk´s 256 TB NVMe SSD is engineered specifically to optimize high-throughput, data-intensive workflows, including data ingestion, preprocessing and fast-moving Artificial Intelligence data lakes. System architects can leverage the drive´s high-speed performance and leading power efficiency, helping to address operational costs and total cost of ownership in hyperscale cloud environments.

This launch signals Sandisk´s commitment to growing enterprise flash storage in sync with evolving data and Artificial Intelligence requirements. By delivering a purpose-built platform capable of supporting vast data repositories and intensive workloads, Sandisk underscores the central role of advanced NAND technologies in tomorrow´s digital infrastructure. The UltraQLC platform, showcased by this exceptional 256 TB SSD, is set to shape the future of enterprise storage, offering both extreme capacity and finely tuned performance for the world´s biggest data challenges.

63

Impact Score

Samsung to supply half of NVIDIA’s SOCAMM2 modules in 2026

Hankyng reports Samsung Electronics has secured a deal to supply half of NVIDIA’s SOCAMM2 modules in 2026 for the Vera Rubin Superchip, which pairs two ‘Rubin’ Artificial Intelligence GPUs with one ‘Vera’ CPU and moves from hardwired memory to DDR5 SOCAMM2 modules.

NVIDIA announces CUDA Tile in CUDA 13.1

CUDA 13.1 introduces CUDA Tile, a virtual instruction set for tile-based parallel programming that raises the programming abstraction above SIMT and abstracts tensor cores to support current and future tensor core architectures. The change targets workloads including Artificial Intelligence where tensors are a fundamental data type.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.