With the spread of agentic AI and the increasing demand for on-site data processing, building low-latency, power-efficient edge AI infrastructure is emerging as a challenge across industries. In this context, Supermicro has expanded its portfolio targeting diverse sectors such as retail, manufacturing, physical security, and logistics by successively unveiling edge AI systems equipped with Intel's latest processors and GPUs.


On the 25th, Supermicro unveiled a new edge AI system based on Intel Core Ultra Series 3 processors, Intel Core Series 2 processors, and Intel Arc Pro B Series GPUs.

It was announced that this portfolio expansion encompasses a variety of form factors ranging from compact fanless systems to 1U rackmount servers and mini towers, aiming to simultaneously boost AI inference performance and Total Cost of Ownership (TCO) efficiency.

There are three major systems unveiled this time.

The fanless SYS-E103-14P is a DIN rail-mounted edge platform equipped with an Intel Core Ultra Series 3 processor, providing up to 180 TOPS of AI performance through its integrated GPU and NPU.

The company explained that it supports up to 128GB of DDR5 memory and has an operating temperature range of 0°C to 45°C, making it suitable for computer vision and industrial automation workloads.

The slim AI mini tower SYS-521AD-LN2 is equipped with an Intel Core Series 2 processor.

It features up to 12 P-cores and up to 64GB of DDR5 memory, and supports both Intel Arc Pro B50 GPUs and NVIDIA RTX Pro Blackwell 2000 GPUs, making it suitable for local AI inference, model development, and fine-tuning in office and edge environments.

The existing 1U SYS-111AD-WN2R and compact SYS-E300-13AD5 have also received updates adding support for Intel Core Series 2 processors.

Supermicro has added support for Intel Arc Pro B-series GPUs across its edge AI server portfolio.

The main specifications are as follows.

The Arc Pro B70 supports high-throughput AI pipelines with up to 367 TOPS and up to 32GB VRAM. The Arc Pro B60 handles multi-GPU scalability and large-scale AI workload processing with up to 197 TOPS. The low-power Arc Pro B50 is suitable for space-constrained edge environments with up to 170 TOPS.

“Enterprises need edge infrastructure that can provide real-time inference, low-latency performance, and power efficiency close to where data is generated,” said Mory Lin, Vice President of IoT, Embedded, and Edge Computing at Supermicro. “The latest Intel-based edge systems and the DCBBS portfolio enable customers to flexibly deploy and scale AI workloads across various edge environments while efficiently managing costs.”