
OCP ORV3 2-OU direct-to-chip liquid-cooled system with 8x NVIDIA HGX B300 SXM GPUs; up to 144 GPUs per rack for hyperscale AI factories.
The Supermicro GPU SuperServer SYS-222GS-NB3OT-ALC 2-OU OCP Liquid-Cooled HGX B300 8-GPU is an enterprise-grade server from Supermicro, engineered for datacenter, virtualization, and AI/HPC workloads. It is built on Dual Intel Xeon 6 (X14 platform) in a dual-socket configuration within a 2-OU (OCP ORV3) chassis.
Storage is handled by 8x front hot-swap E1.S NVMe + 2x M.2 NVMe (RAID via S3808N).
Expansion includes 2x PCIe 5.0 x16 FHHL with SXM GPU support. Networking is provided by 8x NVIDIA ConnectX-8 SuperNICs, up to 800Gb/s. Power is delivered by Power via busbar (48-52V DC, OCP rack), with liquid cooling.
Typical deployments include AI/Deep Learning Training and Inference, Large Language Model (LLM) and Generative AI, High Performance Computing (HPC), Scientific Research, Conversational AI, Business Intelligence & Analytics, Drug Discovery, Finance Services and Fraud Detection, Autonomous Vehicle Technologies.
Configure and compare pricing for the Supermicro GPU SuperServer SYS-222GS-NB3OT-ALC 2-OU OCP Liquid-Cooled HGX B300 8-GPU from verified vendors on GPU Vendors. Request a custom quote with full specifications, build-to-order options, and delivery lead times.
| CPU | Dual Intel Xeon 6 (X14) |
| GPUs | 8x NVIDIA HGX B300 SXM (288GB) |
| Interconnect | NVLink with NVSwitch |
| Drive bays | 8x E1.S NVMe + 2x M.2 NVMe (RAID via S3808N) |
| Fabric | 8x NVIDIA ConnectX-8 SuperNICs up to 800Gb/s |
| PCIe | 2x PCIe 5.0 x16 FHHL |
| Type | Direct-to-Chip (D2C) cold-plate liquid cooling |
| Input | Busbar 48-52V DC (OCP rack) |
| Form factor | 2-OU OCP ORV3 |
| Dimensions | 537mm W x 805mm D |
| Net weight | 70 kg |
We'll email you the moment it happens. No account needed.
Free price-tracking. Unsubscribe anytime.
Comments
Smart picks based on Supermicro GPU SuperServer SYS-222GS-NB3OT-ALC 2-OU OCP Liquid-Cooled HGX B300 8-GPU — upgrades and comparable options.


Scale up for bigger workloads

Scale up for bigger workloads

Scale up for bigger workloads

Comparable system

Comparable system

Comparable system

Comparable system
Earn More