Uvation MarketplaceMarketplace

NVIDIA HGX Rubin NVL8 AI Platform with 8 Rubin GPUs - Vera x 86 CPU Configuration

An OEM-integrated eight-GPU platform with NVIDIA Rubin SXM GPUs and sixth-generation NVLink, available with a Vera CPU or x86 CPU.

8×
NVIDIA Rubin SXM GPUs per platform
2.3 TB
HBM4 GPU memory per platform
400 PFLOPS
NVFP4 inference performance per platform

Vera or x86 CPU, Eight Rubin GPUs

HGX Rubin NVL8 pairs eight Rubin SXM GPUs with either a single NVIDIA Vera CPU or an OEM-selected x86 CPU for agentic AI, analytics and HPC.

Single Vera CPU With 88 Olympus Cores

The Vera configuration adds one NVIDIA Vera CPU with 88 custom Olympus cores and 1.5 TB of LPDDR5X at 1.2 TB/s.

x86 Option Defined by Your OEM

The x86 configuration pairs the GPU baseboard with an OEM-selected x86 CPU and OEM-defined system memory.

130 TB/s of HBM4 Memory Bandwidth

Eight Rubin SXM GPUs combine 2.3 TB of HBM4 memory with 130 TB/s of aggregate bandwidth for data-hungry models.

Eight Rubin GPUs on
One NVLink Baseboard

Sixth-generation NVLink and 28.8 TB/s of NVLink Switch bandwidth connect all eight Rubin SXM GPUs on one baseboard.

NVIDIA HGX Rubin NVL8 AI Platform with 8 Rubin GPUs - Vera or x86 CPU Configuration rack architecture
8×
Rubin SXM GPUs
28.8 TB/s
NVLink Switch bandwidth
130 TB/s
HBM4 bandwidth
88
Vera Olympus cores

Workloads

Designed for agentic AI, data analytics and high-performance computing, with the CPU platform chosen to fit each deployment.

Agentic AI Inference

Up to 400 PFLOPS of sparse NVFP4 inference across eight Rubin GPUs keeps agentic AI responses fast.

Data Analytics at GPU Speed

2.3 TB of HBM4 at 130 TB/s, plus 1.5 TB of Vera LPDDR5X, keeps large datasets close to compute.

High-Performance Computing

Sixth-generation NVLink and 1.6 TB/s of networking bandwidth support simulation and HPC jobs across the platform and beyond.

Engineered to dominate

Eight GPUs, Two CPU Paths

HGX Rubin NVL8 combines Rubin GPU compute, HBM4 bandwidth and sixth-generation NVLink in an eight-GPU baseboard that OEMs build into complete systems.

Exploded view of the NVIDIA HGX Rubin NVL8 AI Platform with 8 Rubin GPUs - Vera or x86 CPU Configuration
400 PFLOPS
NVFP4 inference (sparse)
1.5 TB
Vera LPDDR5X memory
1.2 TB/s
Vera memory bandwidth
1.6 TB/s
Networking bandwidth
Exploded view of the NVIDIA HGX Rubin NVL8 AI Platform with 8 Rubin GPUs - Vera or x86 CPU Configuration
400 PFLOPS
NVFP4 inference (sparse)
1.5 TB
Vera LPDDR5X memory
1.2 TB/s
Vera memory bandwidth
1.6 TB/s
Networking bandwidth

The rack, itemized

SUBSYSTEM

SPECIFICATION / CONFIGURATION DETAILS

Processor

1 × NVIDIA Vera 88-core CPU or OEM-selected x86 CPU

Operating System

N/A

Chipset

NVIDIA HGX Rubin NVL8 8-GPU SXM baseboard

Accelerators / GPU

8× NVIDIA Rubin SXM GPUs; sixth-generation NVLink

Memory

2.3 TB HBM4 GPU memory; Vera configuration adds 1.5 TB LPDDR5X CPU memory; x86 memory is OEM-defined

Primary Storage

N/A

Secondary Storage

N/A

Security

N/A

Chassis

Eight-GPU SXM baseboard; chassis selected by OEM

Power Supply

N/A

Ports

N/A

Slots / PCIe Architecture

6th-generation NVLink bridge; NVLink-C2C CPU interconnect

Bezel

N/A

Form Factor

Module for liquid-cooled NVIDIA MGX modular servers—not a complete server

Network Card

N/A

Warranty

N/A

NVIDIA HGX Rubin NVL8 AI Platform with 8 Rubin GPUs - Vera or x86 CPU Configuration power and networking

Bandwidth for Every Layer
of the Platform

28.8 TB/s NVLink Switch bandwidth

Sixth-generation NVLink Switch bandwidth of 28.8 TB/s moves data quickly between all eight Rubin GPUs.

1.6 TB/s networking bandwidth

The platform specifies 1.6 TB/s of networking bandwidth

1.2 TB/s Vera memory bandwidth

1.5 TB of LPDDR5X delivers 1.2 TB/s of CPU memory bandwidth.

Nuclear-Ready, Plug and Play AI Factories

Bypass grid constraints with self-contained, high-density inference ready infrastructure directly adjacent to clean, base load nuclear power.

Global Deployment Engine

Any Infra, Anywhere in the world.

Global Shipping and Deployment Capability treats advanced infrastructure as mobile, modular freight.

Packaged complex computing nodes, cooling manifolds, and power distribution units into certified, standard form factors.

360° Service Ecosystem

Unifying Legacy with Next-Gen Compute.

Vertically integrated service ecosystem that serves as an operational bridge between legacy environments and next-generation, AI-native infrastructure.

Finance and Leasing

Capital-as-a-Service for High-Density Compute

Convert upfront capital expenditures (CapEx) into predictable operating expenses (OpEx).

Uvation allows enterprises and sovereign entities to deploy multi-megawatt AI factories and advanced GPU clusters without liquidity constraint.