Supermicro GPU SuperServer 751GE-TNRT-NV1

The Supermicro 751GE-TNRT-NV1 is engineered for high-performance computing, making it ideal for tasks like engineering, scientific research, and AI/deep learning training. Its innovative closed-loop liquid cooling system efficiently manages heat for both CPUs and GPUs. The tower features cold and hot tubes, cold plates for the CPU, and a radiator behind the front panel to disperse heat. It supports up to 4x double-width liquid-cooled GPUs, including the NVIDIA PCIe A100, with a PCIe 4.0 x16 connection and an optional NVIDIA NVLink Bridge for added performance. Additionally, the chassis includes 2x liquid level viewing windows and dedicated inlet and outlet ports at the rear.

The front top of the GPU SuperServer 751GE-TNRT-NV1 features 8x 2.5” hot-swap drive bays, accommodating NVMe, SAS, or SATA storage configurations. Note that SAS drives may require additional storage controllers or cables. For enhanced storage capacity, or to host the operating system, the server board includes 2x M.2 NVMe slots, compatible with 2280 or 22110 form factor cards. 2x SATA Disk on Module (DOM) ports provide more options for booting the system or for additional storage. If installing NVMe drives, optional Virtual RAID on CPU or VROC is supported. For networking controllers, storage controllers or GPU acceleration, the system is equipped with 7x PCIe 5.0 x16 FHFL slots, conveniently accessible from the back panel. Powering this setup are dual redundant 2200W Titanium Level (96%) power supplies.

To manage the 751GE-TNRT-NV1, administrators can use Supermicro’s specialized software tools, such as SuperCloud Composer, Supermicro Server Manager, SuperServer Automation Assistant, and Supermicro SuperDoctor 5. SuperDoctor 5 is user-friendly, offering both command-line and web interfaces for Windows and Linux. It keeps track of important system metrics like CPU temperature, voltage levels, power consumption, and fan speed. Plus, it sends alerts through email or Simple Network Management Protocol (SNMP) if anything goes wrong, helping users stay informed about their system’s health.