Super Micro Computer, Inc. (SMCI), a leader in IT solutions for AI, cloud, storage, and 5G/edge applications, has unveiled a new, versatile high-density infrastructure platform tailored for AI inferencing at the network edge. As organizations increasingly adopt complex large language models (LLMs) into their operations, the demand for advanced hardware capable of processing large volumes of data with minimal latency has never been greater. Supermicro's latest offering is a game-changer, providing a robust system that supports up to 10 double-width GPUs, all while fitting within a standard air-cooled environment.
The newly launched 3U EDGE AI Inferencing System, known as SYS-322GB-NR, is designed to meet the needs of modern data processing. It features two powerful Intel® Xeon® 6900 processors equipped with performance cores, along with 8800 MT/s MRDIMM memory and up to 20 PCIe 5.0 expansion slots. This combination allows for flexibility in configuration, enabling users to incorporate various single or double-width GPUs, high-performance I/O options, or other add-on cards as needed.
Charles Liang, president and CEO of Supermicro, emphasized the system's capabilities: “Owing to the system's optimized thermal design, Supermicro can deliver all this performance in a high-density 3U 20 PCIe system with 256 cores that can be deployed in edge data centers.” He highlighted that the system is specifically designed to handle AI workloads, enabling users to run LLM-based applications on-premises, close to the data source, thereby minimizing latency.
One prominent application of the SYS-322GB-NR system is within the manufacturing sector. By deploying this AI infrastructure on-site in automated production environments, businesses can process real-time data from cameras and sensors without the need to transfer data to remote locations. This capability not only reduces networking demands but also significantly enhances response times, making operations more efficient.
Additionally, the system excels in large-scale control rooms, where its AI accelerator cards can be complemented by multi-display cards to support up to 64 independent displays. This versatility makes it suitable for various environments, from monitoring operations to facilitating complex decision-making processes.
The SYS-322GB-NR will be on display at Supermicro’s booth 518 during the Mobile World Congress (MWC) in Las Vegas, taking place from October 8-10. This event offers a platform for Supermicro to demonstrate its innovative technology alongside other advanced systems that integrate NVIDIA, AMD, and Intel Xeon processors.
- SYS-222HE-FTN: A dual Intel Xeon 6 processor system designed to bring data center performance to the telco edge, featuring a compact 2U, short-depth form factor with front I/O access.
- SYS-212B-FN2T: This 2U system is tailored for AI applications in telco and edge deployments, housing a single Intel Xeon 6700 series processor with E-cores and GPU support.
- SYS-E403-14B-FRN2T: A wall-mountable edge device that integrates the Intel Xeon 6700 series processor with E-cores and GPU capabilities, ideal for remote environments.
- AS-1115S-FDWTRT: A 1U NEBS-compliant system utilizing the AMD EPYC 8004 Series processor, providing Telco performance for various workloads.