Tech Gaming Report

Latest Tech & Gaming News

Samsung: UFS 4.0 geht in Serie und Memory-semantic SSD geplant

Samsung: UFS 4.0 enters series production and memory semantic SSD planned

Image: Samsung

Samsung’s press release on the Flash Memory Summit contains little concrete and much of what is already known. The previously introduced new generation of UFS 4.0 smartphone storage is scheduled to go into serial production this month. What exactly is behind the new “memory semantic SSD” remains unclear.

UFS 4.0 and “new” enterprise SSDs

Samsung had already presented in May the new storage for smartphones UFS 4.0 with up to 4,200 MB/s and 1 TB. Now follows the announcement that the memory will go into series production this month. Samsung is expected to implement UFS 4.0 in new flagship smartphones such as Galaxy Z Fold 4 and Galaxy Z Flip 4 among others. Both are expected for the Samsung Galaxy Unpacked event on August 10, 2022.

Samsung is also announcing the availability of the PM1743 and PM1653 enterprise SSDs. Both are also already known, the fast PM1743 with 13 GB/s via PCIe 5.0 was already presented at CES 2022 in January. Much further back is the announcement of the PM1653 with SAS 24G, which Samsung had already presented in April 2021.

“Semantic memory SSD” without details

On the other hand, the “semantic memory SSD” seems innovative, which according to Samsung “combine the advantages of storage and DRAM memoryHowever, Samsung keeps a low profile with details or even specific key data. The only thing that is clear is that both NAND flash and DRAM are used; As is well known, Samsung manufactures both types of memory. It is said that the (Large?) DRAM cache, along with Compute Express Link (CXL), provides significant performance gains over conventional SSDs when accessing small blocks of data, which is important in emerging AI workloads and machine learning.

Leveraging Compute Express Link (CXL) interconnect technology and an embedded DRAM cache, semantic memory SSDs can achieve up to a 20x improvement in both random read speed and latency when used in AI and ML applications. . Optimized to read and write small sized chunks of data at much faster speeds, Samsung Semantic Memory SSDs will be ideal for the growing number of AI and ML workloads that require very fast processing of smaller data sets. .


Petabyte storage for servers or the collection of metadata about SSDs and their components with a new telemetry technique that centers data”identify and prevent potential problems early on” Should be able to.

So far there have been no major announcements, while SK Hynix and YMTC, for example, have introduced their new 3D NAND generations. The Flash Memory Summit 2022 topic page offers more innovations from the memory fair.