8U-CLASS / AI SERVER EXTERIOR CONCEPT
ZXO S8
A considered form
for demanding AI.
An 8U-class exterior concept targeting an NVIDIA HGX H200 8-GPU platform.
Explore the 10-view study
Target rack form factor
NVIDIA H200 SXM target
HBM3e per GPU · NVIDIA reference
01 / THE S8 DIRECTION
An exterior with purpose.
A platform with potential.
S8 explores a composed industrial identity for an eight-GPU AI server: a deep rack enclosure, broad mesh surfaces and restrained metallic detail.
The proposed platform is NVIDIA HGX H200 with eight H200 SXM GPUs. The 8U-class enclosure is ZXO’s design target; mechanical fit, cooling, power delivery and complete-system integration still need validation.
Review the platform targets02 / QUIETLY DISTINCTIVE
Black, with depth.
Gold, with restraint.
An understated material palette gives the S8 study a clear visual identity, from the full enclosure to its smallest exterior details.


A visual language built around generous intake areas, an organized front and a strong horizontal rhythm.
Final airflow, service access and structural details are part of the engineering development brief.03 / TARGET CONFIGURATION
The proposed
HGX H200 foundation.
ZXO enclosure targets and NVIDIA platform references, stated separately from a validated system specification.
NVIDIA HGX H200
8-GPU
H200 SXM GPUs use 141 GB of HBM3e memory per GPU. GPU power is configurable, up to 700 W per GPU.
NVIDIA H200 specifications- Target form factor
- 8U-class rack server
- Target platform
- NVIDIA HGX H200 8-GPU
- Target GPUs
- 8 × NVIDIA H200 SXM
- GPU memory reference
- 141 GB HBM3e per GPU
- GPU power reference
- Up to 700 W per GPU, configurable
- Development status
- Exterior concept; engineering validation pending
8U-class is the ZXO enclosure target, not a standard NVIDIA HGX H200 system height. GPU power is not total system power. CPU, storage, networking, cooling, electrical requirements, pricing and availability are not finalized.
Sources: NVIDIA H200 ↗ · NVIDIA HGX reference architecture ↗

04 / THE WIDER ENVIRONMENT
Think beyond
the enclosure.
Workload, rack space, site power and cooling all shape an AI infrastructure project. The conversation starts with your environment.
Illustrative deployment scene · Not an installed customer system05 / THE COMPLETE DESIGN STUDY
Ten views.
A complete perspective.
From studio views to material details and illustrative deployment scenes. Select any image to enlarge it or open the original artwork.










All ten images are AI-generated exterior-concept visualizations. Rear service zones and interfaces are illustrative, not finalized rear I/O. These images are not CAD projections, validated assemblies or actual deployment photographs.
06 / YOUR NEXT AI PROJECT
Start with
what you need to run.
Share your applications, model requirements, deployment location, available power and cooling, project quantity and timing.
An inquiry starts a discussion. Final configuration, sourcing and availability require confirmation.