Samsung is preparing to usher in the era of supercomputers!

Samsung, one of the world’s largest semiconductor memory chip manufacturers, introduced its new CXL (Compute Express Link) memory modules and HBM3E memory at the event held today. Let’s take a look at the details together.

Today the company introduced new memory modules and HBM3E memory. These products will be used in cloud servers and supercomputers for artificial intelligence and other high-performance computing needs.

At the Memory Con 2024 fair held at the Santa Clara Computer History Museum in Silicon Valley, Samsung announced CXL Memory Module – Box (CMM-B), CXL Memory Module – DRAM (CMM-D), CXL Memory Module Hybrid Layered Memory (CMM-H). TM) and introduced HBM3E 12H memory.

Samsung introduced HBM3E memory

Samsung introduced HBM3E memory

Samsung HBM3E memory, which has high capacity for faster artificial intelligence training and inference, was introduced. Here are the details!

By Broadcom in collaboration with VMWare, Samsung introduced the Peaberry project. This will be the world’s first FPGA-based layered memory for hypervisors. Called CXL Memory Module Hybrid Tiered Memory (CMM-H TM), this memory will be a hybrid solution that combines DRAM and NAND flash memory in an Add-In Card (AIC) format and will be able to tackle memory management challenges, increase performance, optimize throughput times and will reduce the total cost.

Samsung’s HBM3E 12H memory, the world’s first 12-stack HBM3E DRAM memory chip offering the highest capacity achieved using HBM technology, was also exhibited at Memcon 2024. HBM3E 12H memory modules feature Samsung’s advanced thermal compression conductive film (TC NCF) technology, which increases vertical density by more than 20% compared to its previous version. Samsung plans to start mass production for the HBM3E 12H in the first half of this year.

source site-28