SUPERMICRO SYS-120H-TNR 1U Rackmount Server Barebone,Dual Socket P+ (LGA-4189) 3rd Gen Intel® Xeon® Scalable Processors.

The SUPERMICRO SYS-120H-TNR is a 1U dual-socket server barebone for enterprise, cloud, and virtualization deployments that need support for two 3rd Gen Intel Xeon Scalable processors, 32 memory slots, and three PCIe 4.0 x16 slots; CPUs and memory are not included.

Recommended For

• Data-center teams consolidating virtualization or cloud workloads into a 1U rack server, with room to configure dual Xeon processors and large ECC memory capacity.
• Storage and infrastructure teams building fast shared-storage or software-defined-storage nodes that need front-access hot-swap 2.5-inch drives and optional NVMe, SAS, or SATA configurations.
• IT teams running inference or machine-learning workloads that want a rack-dense compute platform with support for up to two double-width GPUs or accelerators.
• Enterprise infrastructure teams that need a configurable server for high-memory workloads, with many DIMM slots and optional networking through an OCP 3.0-compatible AIOM slot.

Real-World Use Scenarios

• When an IT team wants to consolidate virtual machines in a rack, it can configure this server as a dual-processor host with substantial memory capacity; the result is a single 1U platform suited to running multiple virtualized services, depending on the installed CPUs, memory, and hypervisor.
• For a software-defined storage deployment, an operator can combine hot-swap drive bays with an appropriate storage controller and networking option; this supports a rack-based storage node where drives can be serviced without removing the server from operation.
• When a team needs to run AI inference close to its existing rack infrastructure, it can configure the system with supported accelerator cards; this provides a compact inference host, with suitable performance depending on the selected accelerators and workload.
• For cloud or enterprise service hosting, administrators can use the server as a dense rack node for compute services and internal applications; its 1U format helps fit more configured systems into limited rack space.
• When an application relies on large in-memory datasets, a team can configure the platform with high-capacity memory; this can reduce reliance on slower storage for data that fits in memory, with the actual benefit depending on the application and installed memory.

Spec Summary