Delivering AI Performance to the Intelligent Edge, Powered by Intel
The strategic partnership between Supermicro and Intel has been a vital component in Supermicro’s edge AI portfolio for years, resulting in a broad range of powerful, performance-efficient systems designed for space-constrained and remote environments.
Intel’s ground-breaking CPU technology delivers unmatched compute and AI performance at the edge, enabling businesses to adopt and scale data-driven applications with speed, performance, and trust. On top of that, Intel’s OpenVINO framework helps developers accelerate AI inference workloads and efficiently deploy scalable solutions across hardware platforms.
Intel’s edge AI portfolio consists of specialized CPU, GPU, and NPU units, ensuring an exact match to any use case’s workload and requirements.
CPU-Integrated AI Acceleration
Intel’s AI-enabled processors are built for demanding edge workloads, including data inferencing. Advanced Vector Extensions (AVX2) enable efficient parallel data processing, allowing small language models and similar AI workloads to run directly on the CPU, reducing the need for external processing.
Intel Core Ultra processors extend this capability with integrated GPU and NPU acceleration. The integrated GPU shares its memory with the CPU and excels at processing graphics and accelerated workloads. The NPU handles low-power AI tasks independently, freeing CPU resources and improving overall efficiency.
This performance-efficient architecture makes Intel processors ideal for AI deployments across retail, manufacturing, healthcare, and other edge environments.
Reliable, Scalable AI Performance
Built for professional media creation and edge AI, Intel Arc Pro B50 GPU and Intel Arc Pro B60 GPU deliver powerful AI acceleration based on Intel’s Xe2 architecture.
With up to 24GB VRAM, these GPU cards support demanding workloads such as generative AI and local model inference. The compact B50 (70W) fits edge devices and small systems with up to 170 TOPS, while the B60 delivers higher compute performance (up to 197 TOPS), greater memory bandwidth, and multi-GPU scalability for larger models.
Industry-standard APIs, ISV certifications, and PCIe 5 connectivity make it easy to deploy AI workloads on-premises or at the network edge—reducing latency, improving data control, and minimizing cloud dependency.
Edge AI for Vertical Applications
Manufacturing
Edge AI inferencing in manufacturing enables real-time quality inspection, predictive maintenance, and process optimization on factory floors. Ruggedized, low-latency hardware with GPU acceleration supports vision workloads, while energy-efficient designs handle harsh environments and operating temperatures to deliver continuous operation without cloud dependence.
Retail
Edge AI inferencing in retail powers in-store analytics, loss prevention, and personalized promotions with instant responses. Compact, cost-effective hardware with integrated accelerators processes video and sensor data locally, balancing privacy, bandwidth limits, and scalability across distributed large diverse store environments.
Healthcare
Edge AI inferencing in healthcare enables diagnostics, patient monitoring, and imaging analysis at the point of care. Purpose-built hardware with specialized accelerators and low-power operation supports secure data processing, ensuring reliability, privacy, and real-time decision-making in clinical and remote settings remote settings.
Transportation
Edge AI inferencing in transportation enables real-time traffic management, autonomous navigation, and fleet monitoring at the network edge. High-performance, low-latency hardware with AI accelerators processes sensor and video streams in real-time, while rugged, compact designs ensure reliability across vehicles, infrastructure, and dynamic operating environments.
Fanless Edge AI System with Intel® Core™ Ultra Series 3 Processor
Compact Fanless Industrial Computer Designed for Advanced Edge AI: SYS-E103-14P Datasheet
- CPU: Intel® Core™ Ultra X7 Processor 368H
- Graphics: Intel Arc™ B390
- Memory 2× DDR5 SO-DIMM slots (2 channels), up to 128 GB DDR5-5600 non-ECC (1DPC), 1.1V
- Storage M.2-only storage with 1× PCIe 5.0 x4 M-Key (2242/2280), 1× PCIe 4.0 x2 M-Key (2280), plus B-Key (w/ nano-SIM) and E-Key slots; supports up to two NVMe 2280 SSDs Wireless Network M.2 E-Key 2230 slot (PCIe 4.0 ×1/USB2.0) for optional Wi-Fi/BT modules; M.2 B-Key 3052 with nano-SIM for optional WWAN/LTE/5G
- LAN 2× 2.5GbE (Intel® Ethernet Controller I226-IT, optional PoE) + 1× 1GbE (Intel® Ethernet Connection I219-LM)
- OS Windows (IoT/Pro) and Linux
- I/O Options (Input/Output): 2x USB 2.0 Type-A, 2x USB 3.2 Type-A 2x HDMI 2.1, 2x COM RS232/RS422/RS485, 1x SMBus 8-bit GPIO
Intel Products and Technologies: Accelerating AI and Analytics at the Edge
Low-Profile, Mini-1U System, up to 65W TDP CPU (Air Cooled)
Ideal Low-Profile, Enterprise Edge Server for AI Inferencing: SYS-E300-14AR Datasheet
- CPU: Single Intel® Core™ Ultra 9/7/5 processor (Series 2)
- Graphics: (1) low-profile PCIe 5.0x16 GPU
- Memory: (2) DIMM slots, Max Memory (1DPC): Up to 96GB 6400MT/s ECC/non-ECC DDR5 SO-DIMM
- Storage:
(1) Internal fixed 2.5” SATA drive bay
(1) M.2 PCIe 5.0x4NVMe slot (M-key 2280) - Wireless Network: Available through an add-on module
- LAN:
(1) RJ45 1 GbE Dedicated BMC LAN port
(2) RJ45 10 GbE LAN ports (Intel® X550)
(2) RJ45 2.5 GbE LAN ports (Intel® I226-LM) - OS Windows 11, RHEL 10.0, and Ubuntu 24.04
Intel Products and Technologies: Accelerating AI and Analytics at the Edge
1U Short-Depth Rackmount Server for Edge Compute
Compact Server with Intel® Xeon® 6 Processors & 2 Full PCIe Slots: SYS-112B-FWT Datasheet
- 1U short-depth, air-cooled rackmount server with Intel® Xeon® 6 Scalable processors supporting 85W - 250W TCP.
- Designed with redundant AC power for IoT and edge compute applications such as Edge AI and Telco virtual RAN / open RAN
- The Supports 8 DIMM slots with a maximum system memory up to 2TB 5600MT/s ECC DDR5 RDIMM
- Storage:
(2) internal fixed 2.5” SATA drive bays and
optional (1) M.2 PCIe 3.0x2 NVMe slot (M-key 2280/22110) - Supports front I/O networking with
(2) RJ45 10GbE LAN ports (Intel® Ethernet Controller X550-AT2) and
(1) RJ45 1GbE dedicated BMC LAN port for IPMI remote management - Supports multiple PCIe configurations:
(2) PCIe 5.0x16 FHFL slots and (1) PCIe 5.0x16 Low Profile slot
Intel Products and Technologies: Accelerating AI and Analytics at the Edge
Powerful Compact Enterprise Server for Edge AI and Compute
Enterprise Edge Server with Multiple GPUs for AI Inferencing: SYS-E403-13E-FRN2T Datasheet
- CPU: Single Socket E (LGA-4677), 5th Gen Intel® Xeon® / 4th Gen Intel® Xeon® Scalable processors
- Graphics: Up to (1) double-width or (2) single-width GPUs
- Memory: (8) DIMM slots, max memory (1DPC): Up to 2TB 5600MT/s ECC DDR5 RDIMM
- Storage: Total (4) bays.
(2) front hot-swap 2.5" NVMe drive bays and
(2) internal fixed 2.5" SATA* drive bays - Wireless Network: Available through an add-on module
- LAN:
(1) RJ45 1 GbE Dedicated BMC LAN port and
(2) RJ45 10 GbE LAN ports (Intel® Ethernet Controller X550-AT2) - OS: Windows Server 2022, RHEL 9.2, Oracle Linux 9.2, Ubuntu 22.04
