DDR5 DRAM-based reminiscence module is designed to satisfy the high-performance calls for of data-intensive purposes together with AI and HPC
CXL interface allows reminiscence capability to scale to the terabyte stage and considerably reduces system latency
Samsung Electronics, the world chief in superior reminiscence know-how, right now unveiled the {industry}’s first reminiscence module supporting the brand new Compute Express Link (CXL) interconnect commonplace. Integrated with Samsung’s Double Data Rate 5 (DDR5) know-how, this CXL-based module will allow server programs to considerably scale reminiscence capability and bandwidth, accelerating synthetic intelligence (AI) and high-performance computing (HPC) workloads in knowledge facilities.
The rise of AI and massive knowledge has been fueling the development towards heterogeneous computing, the place a number of processors work in parallel to course of huge volumes of information. CXL—an open, industry-supported interconnect based mostly on the PCI Express (PCIe) 5.zero interface—allows high-speed, low latency communication between the host processor and gadgets comparable to accelerators, reminiscence buffers and sensible I/O gadgets, whereas increasing reminiscence capability and bandwidth nicely past what is feasible right now. Samsung has been collaborating with a number of knowledge heart, server and chipset producers to develop next-generation interface know-how because the CXL consortium was shaped in 2019.
“This is the industry’s first DRAM-based memory solution that runs on the CXL interface, which will play a critical role in serving data-intensive applications including AI and machine learning in data centers as well as cloud environments,” mentioned Cheolmin Park, vp of the Memory Product Planning Team at Samsung Electronics. “Samsung will continue to raise the bar with memory interface innovation and capacity scaling to help our customers, and the industry at-large, better manage the demands of larger, more complex, real-time workloads that are key to AI and the data centers of tomorrow.”
Dr. Debendra Das Sharma, Intel Fellow and Director of I/O Technology and Standards at Intel mentioned, “Data center architecture is rapidly evolving to support the growing demand and workloads for AI and ML, and CXL memory is expected to expand the use of memory to a new level. We continue to work with industry companies such as Samsung to develop a robust memory ecosystem around CXL.”
Dan McNamara, senior vp and basic supervisor, Server Business Unit, AMD, added, “AMD is committed to driving the next generation of performance in cloud and enterprise computing. Memory research is a critical piece to unlocking this performance, and we are excited to work with Samsung to deliver advanced interconnect technology to our data center customers.”
Unlike typical DDR-based reminiscence, which has restricted reminiscence channels, Samsung’s CXL-enabled DDR5 module can scale reminiscence capability to the terabyte stage, whereas dramatically decreasing system latency brought on by reminiscence caching.
In addition to CXL {hardware} innovation, Samsung has integrated a number of controller and software program applied sciences like reminiscence mapping, interface changing and error administration, which is able to permit CPUs or GPUs to acknowledge the CXL-based reminiscence and put it to use as the primary reminiscence.
Samsung’s new module has been efficiently validated on next-generation server platforms from Intel, signaling the start of an period for high-bandwidth, low latency CXL-based reminiscence utilizing the most recent DDR5 commonplace. Samsung can be working with knowledge heart and cloud suppliers world wide to raised accommodate the necessity for larger reminiscence capability that will probably be important in dealing with huge knowledge purposes together with in-memory database programs.
As the DDR5-based CXL reminiscence module turns into commercialized, Samsung intends to steer the {industry} in assembly the demand for next-generation high-performance computing applied sciences that depend on expanded reminiscence…
