Supermicro SYS-420GP-TNR Dual Xeon Scalable 4U GPU SuperServer

SuperServer SYS-420GP-TNR: Ultimate 4U GPU Server for AI and High-Performance Computing

Boost your data center capabilities with the Supermicro SuperServer SYS-420GP-TNR, a high-density 4U rackmount GPU system optimized for AI/deep learning training, high-performance computing (HPC), rendering, VDI, and demanding simulations. This dual-socket Intel-based server supports up to 10 double-width PCIe GPUs, including NVIDIA H100, L40, RTX A6000, A100, A40, and Intel Data Center GPU Flex 170, with PCIe 4.0 x16 dual-root interconnects and optional NVIDIA NVLink Bridge for enhanced GPU-to-GPU communication.

Powered by 3rd Gen Intel® Xeon® Scalable processors with up to 40 cores per CPU and 270W TDP (air-cooled), it delivers exceptional processing power. Memory scales to 8TB of 3200MT/s ECC DDR4 RDIMM/LRDIMM across 32 slots, plus up to 8TB Intel Optane Persistent Memory for data-intensive tasks. Storage features 24 front hot-swap 2.5" bays (configurable for NVMe/SATA/SAS), with RAID support and 2 M.2 NVMe slots for boot drives. Expansion includes 12 PCIe 4.0 x16 FHFL slots and AIOM/OCP 3.0 support for flexible networking up to 200GbE via adapters like Intel X550 or Mellanox ConnectX-6.

Engineered for reliability, this AI GPU server includes 4x 2000W Titanium-level redundant power supplies, 8 hot-swap heavy-duty PWM fans for optimal cooling in 10°C to 35°C environments, and advanced thermal monitoring. Security is robust with TPM 2.0, NIST 800-193 compliant Silicon Root of Trust, Secure Boot, cryptographically signed firmware, and system lockdown. The chassis measures 7" H x 17.2" W x 29" D, with a black front and silver body for easy rack integration. NVIDIA-certified for enterprise deployments.

Enhance your HPC and AI infrastructure with the SYS-420GP-TNR—Supermicro's versatile solution for scalable computing. Ideal for enterprises advancing in deep learning and rendering.