Lenovo ThinkSystem SR630 V2 1U GPU Server Triple GPU Support Dual Xeon Gold 5317 128GB AI Inference Platform

Lenovo ThinkSystem SR630 V2 – GPU-Accelerated AI Inference & Edge Computing Server

The Lenovo ThinkSystem SR630 V2 in GPU configuration supports up to 3 single-wide GPUs in a dense 1U chassis, making it one of the most compact GPU-accelerated servers available. Powered by dual Intel Xeon Gold 5317 processors, this platform is optimized for AI inference at the edge, real-time video analytics, and GPU-accelerated virtualization.

Key Features

• Processor: Dual Intel Xeon Gold 5317 – 12 cores, 3.0 GHz base clock for low-latency AI preprocessing and data pipeline
• GPU Capacity: 3x single-wide GPU slots supporting NVIDIA A2 (Tensor Core), NVIDIA T4, NVIDIA L4, or Intel Data Center GPU Flex Series
• Memory: 128GB DDR4-3200 ECC RDIMM with 32 DIMM slots
• Storage: 2x 1.92TB NVMe SSD hot-swap for high-speed model storage and inference data
• Network: OCP 3.0 dual-port 25GbE SFP28 for low-latency data ingestion
• Power: Dual 1800W Titanium PSU for reliable power delivery under full GPU load
• Cooling: Up to 8 N+1 high-performance 40mm dual-rotor fans ensuring thermal stability

Ideal Applications

AI model inference serving (TensorFlow Serving, Triton Inference Server, PyTorch Serve), real-time video analytics and computer vision at the edge, virtual desktop infrastructure with GPU profiles, retail analytics, manufacturing quality inspection AI, and telecommunications edge computing (MEC) with GPU acceleration.

Lenovo XClarity Controller enables remote GPU monitoring and management. 3-year onsite warranty included.