Can AI models grow faster than storage can keep up? An enterprise SSD aims to close that gap with higher speed and lower power.
Samsung Electronics has begun mass production of the PM1763, a PCIe 6.0 enterprise solid-state drive (SSD) designed for artificial intelligence (AI) and high-performance computing (HPC) servers. The storage device can accelerate AI training and inference by delivering higher throughput, improved power efficiency, and enhanced security for next-generation data centres.
As AI models continue to increase in size and complexity, storage performance has become a critical factor in maintaining efficient data movement between processors and accelerators. The PM1763 addresses this challenge with Samsung’s ninth-generation V-NAND flash memory and a newly developed 4 nm controller, enabling higher bandwidth while reducing energy consumption. The company says the SSD has completed validation for next-generation AI platforms, making it suitable for enterprise AI deployments.
The drive is offered in 4 TB, 8 TB, and 16 TB capacities. The 16 TB version delivers sequential read speeds of up to 28,400 MB/s and write speeds of up to 21,900 MB/s, more than doubling the performance of its predecessor. According to Samsung, the SSD can transfer a 40 GB large language model in around 1.4 seconds, helping reduce latency during AI workloads.
The PM1763 is also optimised for liquid-cooled servers through direct-to-chip cooling, allowing sustained performance under continuous high-load operation. In addition, it delivers more than 1.8 times better power efficiency than the previous generation and incorporates post-quantum cryptography alongside the TEE Device Interface Security Protocol to strengthen data protection in virtualised environments.
“Built on industry-leading performance, PM1763 has successfully completed validation for next-generation AI platforms and is well positioned to support evolving AI infrastructure requirements. As AI models continue to grow in size and complexity, PM1763 will serve as a key solution that enables customers to efficiently scale memory capacity and optimize AI operations,” says Jangseok Choi, Vice President and Head of Memory Product Planning at Samsung Electronics.
Click here for the official announcement.
