These edge AI advancements by Supermicro represent a significant step in making AI more accessible and effective.
Supermicro is advancing its AI solution offerings, focusing on edge computing capabilities. This move allows for the application of AI directly where data is generated, like public spaces and retail stores, which enhances responsiveness and decision-making processes. The integration of Supermicro servers with NVIDIA GPUs expands the range of powerful servers available for AI tasks right at the edge.
Supermicro’s president and CEO, Charles Liang, states that their Hyper-E server, powered by dual 5th Gen Intel Xeon processors, supports up to three NVIDIA H100 PCIe GPUs. This configuration provides significant performance for AI applications at the edge, with servers featuring up to 8TB of memory, bringing the processing power typically found in data centers to edge locations.
The company’s developments enable users to process AI data locally at the edge, eliminating the need for data to travel back to cloud centers. This direct processing supports the use of pre-trained AI models, such as those available in NVIDIA AI Enterprise, for more accurate and timely decision-making.
The Hyper-E server model SYS-221HE is designed for optimal performance in edge AI applications, supporting dual socket CPUs and accommodating up to three double-width NVIDIA GPUs. This capability allows for the migration of AI workloads to where data is collected, enhancing the processing and storage of edge data.
Supermicro’s edge server lineup includes various models tailored for different edge computing needs:
- The SYS-111E-FTWR is a high-density edge system, ideal for networking and edge applications.
- The SYS-211E-FRN2T is designed for space-constrained environments and supports up to a 5th Gen Intel Xeon processor.
- The SuperEdge, SYS-211SE-31D, features three independent nodes, each with extensive memory and front I/O, suitable for deployment outside traditional data centers.
- The SYS-E403-13E offers data center-level performance in a compact box PC form factor, ideal for deployment in small spaces.
- The SYS-E403-13E-FRN2T is a compact IoT server that fits an NVIDIA GPU, perfect for distributed AI capabilities at the edge.
These advancements by Supermicro represent a significant step in making AI more accessible and effective in edge environments. By equipping edge servers with substantial computational power, Supermicro is enabling enterprises to tap into AI’s potential without the latency associated with cloud processing.
Engage with StorageReview
Newsletter | YouTube | Podcast iTunes/Spotify | Instagram | Twitter | TikTok | RSS Feed