Triple-Engine AI Architecture

At the core of the SB860R8 is Intel's latest Core™ Ultra 200 Series processor, integrating CPU compute, Intel® Xe graphics, and a dedicated NPU into a single platform. This triple-engine design delivers up to 36 TOPS of combined AI acceleration. Enabling on-device inference without cloud dependency. Alongside up to 24 hybrid cores and support for up to 192 GB of DDR5-5600 memory.

Scalable Expansion for GPU and AI Accelerators

The SB860R8 supports next-generation discrete GPUs and AI accelerator cards through a PCIe Gen5 x16 primary slot and a secondary PCIe Gen4 x4 slot. Storage is equally flexible, with M.2 NVMe for primary drives and four SATA bays supporting RAID 0, 1, 5, and 10 configurations. An optional serial port and dedicated WLAN module bay extend compatibility with industrial peripherals and wireless connectivity requirements.

Quad Display Output

Four independent display outputs. Two HDMI (2.1 and 2.0b) and two DisplayPort 1.4a allow the SB860R8 to drive multi-monitor workstations, operator dashboards, digital signage arrays, and video walls simultaneously.

Enterprise Networking and Storage Protection

Dual Intel® 2.5GbE LAN ports provide high-speed, redundant connectivity for edge server deployments. Combined with hardware RAID and Wake on LAN support, the SB860R8 is ready for 24/7 remote-managed deployments out of the box.

Advanced Thermal Management

An upgraded thermal module with optimized front-intake and side-exhaust airflow delivers targeted cooling for the CPU, GPU, and storage drives independently. Sustaining performance under the continuous heavy loads typical of AI inference and video processing workloads.

Reliability and Security

The SB860R8 integrates Firmware TPM 2.0 and enhanced BIOS reliability features for enterprise-grade stability and security compliance. A 500W 80Plus Gold power supply provides efficient, reliable power headroom for fully configured systems.

Compact Form, Workstation Capability

Measuring just 33.2 × 21.5 × 19.0 cm, the SB860R8 fits into constrained edge environments while accommodating GPU cards up to 280 × 120 × 40 mm. Available as a barebone system, it gives integrators and solution providers the flexibility to specify CPU, memory, storage, and GPU to exact application requirements.