Samsung has initiated the mass production of its latest enterprise solid-state drive, the PM1763, designed specifically to meet the soaring data demands of advanced AI and high-performance computing (HPC) environments. This PCIe 6.0-based SSD sets a new standard in speed and efficiency, supporting the intensive workload requirements of AI platforms like NVIDIA’s Vera Rubin servers.
Equipped with Samsung’s 9th-generation V-NAND technology and a cutting-edge 4-nanometer controller, the PM1763 delivers sequential read speeds up to 28,400 megabytes per second and write speeds reaching 21,900 megabytes per second. These figures represent more than double the performance of its predecessor, the PM1753, enabling rapid transfers such as moving 40-gigabyte large language models in just over one second. This advancement shrinks data latency between processors and accelerators, significantly boosting AI computing efficiency.
The SSD comes in multiple capacities of 4TB, 8TB, and 16TB, with the highest capacity configuration pushing industry-leading performance limits. Additionally, the PM1763 incorporates direct-to-chip (D2C) liquid cooling technology, allowing servers to sustain peak output during prolonged and demanding operations. Its power efficiency also improves by over 80 percent compared to earlier models, which helps lower energy consumption and reduce data center costs.
By delivering a storage solution that balances blistering speeds with thermal and energy management, Samsung’s PM1763 is positioned as a crucial component for next-generation AI infrastructures and data centers aiming to handle rapidly expanding datasets and intensive machine learning workloads.