MSI positions the WS300 as bringing data-center-class AI performance to deskside deployments, aiming to move AI initiatives from experimentation to production.

MSI highlights a strategic collaboration with NVIDIA to bridge centralized performance with distributed AI work, emphasizing speed, scale, and control over data and IP.

Super Micro released seven integrated AI Data Platform solutions built on NVIDIA reference architectures, featuring RTX PRO 6000/4500 Blackwell GPUs, Spectrum-X networking, NVIDIA NIM microservices, and NeMo.

Early buyers include Snowflake, EPRI, Medivis, and research teams at Microsoft and Cornell; units will ship through ASUS, Dell, GIGABYTE, MSI, and Supermicro, with HP joining later; pricing is expected to be six-figure.

Supermicro describes its Server Building Block Solutions and green computing initiative, emphasizing in-house design/manufacture and a broad, configurable product family.

Partners such as Cloudian, DDN, Everpure, IBM, Nutanix, VAST Data, and WEKA will deliver turnkey AI infrastructure for enterprises.

At 20 petaflops, the Station delivers significant performance in a form factor that plugs into a wall outlet, reducing latency and bottlenecks for large models loaded into memory.

WS300 is powered by the NVIDIA Grace Blackwell Ultra Desktop Superchip, supporting up to 748GB of memory and dual 400GbE networking for up to 800Gb/s bandwidth.

The DGX Station combines a Grace CPU and Blackwell Ultra GPU with NVLink-C2C to provide 1.8 TB/s coherent bandwidth between CPU and GPU, enabling a shared memory pool and scalable distributed AI workloads.

The WS300 can serve as a centralized AI compute node for teams while keeping proprietary data and IP under organizational control, extending data-center performance to deskside production-ready AI.

The solutions are showcased at the NVIDIA GTC conference, March 16-19, in Supermicro’s booth.

NVIDIA’s broader GTC 2026 strategy includes Vera Rubin hardware, Vera CPU orchestration, Space Module for orbital data centers, and partnerships across Adobe, automakers, and open frontier model initiatives, positioning NVIDIA as a full-stack AI platform provider.