Uvation MarketplaceMarketplace

High-performance compute
Built for AI

From dedicated training clusters to inference-optimized pools, USP lets you scope GPU density and interconnect topologies.

Five builders.
One accelerated frontier.

The latest NVIDIA platform, delivered through purpose-built systems from the infrastructure leaders you specified.

Nvidia
GB300 NVL72 • DGX Vera Rubin
Dell
PowerEdge XE9780 • XE9712
Supermicro
HGX B300 • Rack-scale GB300
Lenovo
NVIDIA GB300 NVL72
HPE
NVIDIA GB300 NVL72 by HPE

One shortlist.
No legacy systems.

Filter the field by accelerator generation. Select up to three systems for a quick side-by-side view.

Compare products

NVIDIA DGX B300 AI Server

NVIDIA DGX B300 AI Server
Memory37 TB Total Fast Memory; 20 TB GPU Memory
Accelerators72x NVIDIA Blackwell Ultra GPUs
ChipsetNVIDIA Grace Blackwell Ultra Superchips
Processor36x NVIDIA Grace CPUs; 2,592 Arm Neoverse V2 cores

NVIDIA HGX B300

NVIDIA HGX B300
Memory2.1 TB total GPU memory (HBM3e; 288 GB per GPU)
Accelerators8× Blackwell Ultra SXM; 144 sparse / 108 dense PFLOPS FP4; 72 sparse PFLOPS FP8/FP6; 5th-Gen NVLink; 14.4 TB/s aggregate NVLink bandwidth
ChipsetNVIDIA HGX B300 8-GPU SXM baseboard
ProcessorNot specified (standalone baseboard platform; host CPUs provided by system OEMs)

NVIDIA GB300 NVL72

NVIDIA GB300 NVL72
Memory37 TB total fast memory (20 TB GPU HBM3e memory; 17 TB LPDDR5X CPU memory)
Accelerators72× Blackwell Ultra GPUs; 1,440 sparse / 1,080 dense PFLOPS FP4; 720 PFLOPS FP8/FP6; 130 TB/s aggregate NVLink bandwidth
ChipsetNVIDIA GB300 compute tray with 2× GB300 Grace Blackwell Ultra Superchips
Processor36× NVIDIA Grace CPUs; 2,592 Arm Neoverse V2 cores

NVIDIA DGX SuperPOD with DGX B300 Systems — Quantum-X800 InfiniBand, AC Power

NVIDIA DGX SuperPOD with DGX B300 Systems — Quantum-X800 InfiniBand, AC Power
MemoryPer compute node: 2 TB standard, up to 4 TB system memory
Accelerators8× Blackwell Ultra GPUs per node; reference SU: 576 GPUs
ChipsetNA
Processor2 × Intel Xeon 6776P per node

NVIDIA DGX SuperPOD with DGX B300 Systems — Spectrum-4 Ethernet, DC Busbar

NVIDIA DGX SuperPOD with DGX B300 Systems — Spectrum-4 Ethernet, DC Busbar
MemoryPer compute node: 2 TB standard, up to 4 TB system memory
Accelerators8× Blackwell Ultra GPUs per node; reference SU: 512 GPUs
ChipsetNA
Processor2 × Intel Xeon 6776P per node

NVIDIA DGX SuperPOD with DGX B300 Systems

NVIDIA DGX SuperPOD with DGX B300 Systems
MemoryN/A Configurable Product
AcceleratorsN/A Configurable Product
ChipsetN/A Configurable Product
ProcessorN/A

Dell PowerEdge XE9785L AI Server

Dell PowerEdge XE9785L AI Server
Memory24 DDR5 RDIMM slots; up to 6400 MT/s; maximum 6 TB
Accelerators8 × NVIDIA HGX B300 1200 W SXM7 GPUs with NVIDIA NVLink
ChipsetN/A
ProcessorTwo 5th Generation AMD EPYC 9005 Series processors with up to 192 cores per processor

Dell PowerEdge XE9785 AI Server with 8 x NVIDIA HGX B300 270 GB 1100 W SXM6 GPUs

Dell PowerEdge XE9785 AI Server with 8 x NVIDIA HGX B300 270 GB 1100 W SXM6 GPUs
MemoryUp to 24 DDR5 DIMM slots, up to 6400 MT/s, 6 TB maximum
Accelerators8 x NVIDIA HGX B300 270 GB 1100 W SXM6 GPUs, fully interconnected with NVIDIA NVLink technology
ChipsetN/A
ProcessorTwo 5th Generation AMD EPYC 9005 Series processors with up to 192 cores per processor

Dell PowerEdge XE9780L AI Server with 8 x NVIDIA HGX B300 1100W SXM6 GPUs

Dell PowerEdge XE9780L AI Server with 8 x NVIDIA HGX B300 1100W SXM6 GPUs
MemoryUp to 24 DDR5 DIMM slots, up to 6400 MT/s, 6 TB maximum.
Accelerators8 x NVIDIA HGX B300 270 GB 1100 W SXM6 GPUs, fully interconnected with NVIDIA NVLink technology
ChipsetN/A
ProcessorTwo 5th Generation AMD EPYC 9005 Series processors with up to 192 cores per processor

Dell PowerEdge XE9780LAP AI Server

Dell PowerEdge XE9780LAP AI Server
Memory24 DDR5 RDIMM slots; maximum 6 TB
Accelerators8 × NVIDIA HGX B300 NVL8 1100W SXM6 GPUs with NVIDIA NVLink
ChipsetN/A
ProcessorTwo Intel Xeon 6 Processors with up to 128 cores per processor

Dell PowerEdge XE9780 AI Server with 8 NVIDIA HGX B300 NVL8 270 GB 1100 W SXM6 GPUs

Dell PowerEdge XE9780 AI Server with 8 NVIDIA HGX B300 NVL8 270 GB 1100 W SXM6 GPUs
Memory32 DDR5 DIMM slots, up to 6400 MT/s RDIMM 4 TB max (32 x 128 GB)
Accelerators8 NVIDIA HGX B300 NVL8 270 GB 1100 W SXM6 GPUs, fully interconnected with NVIDIA NVLink technology
ChipsetN/A
ProcessorTwo 6th Generation Intel® Xeon® Scalable Processors with up to 86 cores per Processor

Lenovo ThinkSystem SR680a V4 Server

Lenovo ThinkSystem SR680a V4 Server
Memory32 DDR5 DIMMs; up to 4 TB
Accelerators8 x NVIDIA HGX B300 NVL8 GPUs
ChipsetN/A
Processor2× Intel Xeon 6700P; up to 86 cores each

HPE ProLiant Compute XD685 DLC High Power Server

HPE ProLiant Compute XD685 DLC High Power Server
Memory24 DDR5 DIMMs Up to 6.0 TB / 6.1 TB using 256 GB or 128 GB RDIMMs running at speeds up to 6400 MT/s
Accelerators8× NVIDIA B300 SXM; 288 GB per GPU
ChipsetHPE iLO 6 ASIC
Processor2x 5th Generation AMD EPYC Processors

HPE Compute XD690 AI Server

HPE Compute XD690 AI Server
Memory32 DDR5 DIMMs; up to 3 TB
Accelerators8× NVIDIA B300 SXM; 288 GB per GPU
ChipsetN/A
Processor2x Intel® Xeon® 6 processors with P-cores

Supermicro SYS-422GS-NB3RT-ALC GPU Server

Supermicro SYS-422GS-NB3RT-ALC GPU Server
Memory32 DDR5 ECC RDIMM slots; up to 4 TB (at 6400 MT/s) or 8 TB (at 6000 MT/s)
AcceleratorsNVIDIA SXM: HGX B300 8-GPU (288GB HBM3e per GPU, ~2.3 TB total HBM3e)
ChipsetSystem on Chip (SoC)
Processor2× Intel Xeon 6700P (up to 86 cores and 172 threads per processor)

Supermicro SYS-422GS-NB3RT-LCC GPU Server

Supermicro SYS-422GS-NB3RT-LCC GPU Server
Memory32 DIMMs; up to 4 TB or 8 TB, depending on speed
AcceleratorsNVIDIA SXM: HGX B300 8-GPU (288GB)
ChipsetSystem on Chip (SoC)
Processor2× Intel Xeon 6700P; up to 86 cores each

Supermicro AS-4126GS-NB3RT-LCC GPU Server

Supermicro AS-4126GS-NB3RT-LCC GPU Server
Memory24 DIMMs; up to 6 TB
AcceleratorsNVIDIA SXM: HGX B300 8-GPU (288GB)
ChipsetSystem on Chip (SoC)
Processor2× AMD EPYC 9005 or 9004 processors

Supermicro SYS-222GS-NB3OT-ALC GPU Server

Supermicro SYS-222GS-NB3OT-ALC GPU Server
Memory32 DIMMs; up to 4 TB or 8 TB, depending on speed
AcceleratorsNVIDIA SXM: HGX B300 8-GPU (288GB)
ChipsetSystem on Chip (SoC)
Processor2× Intel Xeon 6700P; up to 86 cores each

Supermicro SYS-822GS-NB3RT GPU Server

Supermicro SYS-822GS-NB3RT GPU Server
Memory32 DIMMs; up to 4 TB or 8 TB, depending on speed
AcceleratorsNVIDIA SXM: HGX B300 8-GPU (288GB)
ChipsetSystem on Chip (SoC)
Processor2× Intel Xeon 6700P; up to 86 cores each

Supermicro GB300 NVL72 SuperCluster

Supermicro GB300 NVL72 SuperCluster
MemoryApproximately 21 TB HBM3e plus 17 TB LPDDR5X
Accelerators72× B300 Tensor Core GPUs; 288 GB each
ChipsetGB300 Grace Blackwell Ultra Superchips
Processor36× NVIDIA 72-core Grace ARM Neoverse V2 CPUs

NVIDIA DGX GB300 AI Server

NVIDIA DGX GB300 AI Server
Memory37 TB Total Fast Memory; 20 TB GPU Memory
Accelerators72x NVIDIA Blackwell Ultra GPUs
ChipsetNVIDIA Grace Blackwell Ultra Superchips
Processor36x NVIDIA Grace CPUs; 2,592 Arm Neoverse V2 cores

Dell PowerEdge XE9712 AI Server

Dell PowerEdge XE9712 AI Server
Memory480 GB of LP DDR5 memory with ECC / CPU ; 288 GB HBM3e / GPU
Accelerators4 × NVIDIA Blackwell Ultra GPUs; 900 GB/s NVLink-C2C CPU-to-GPU interconnect
ChipsetNot specified (NA) [1]
ProcessorTwo NVIDIA Grace with 72 cores per processor

Lenovo NVIDIA GB300 NVL72 Rack Scale AI Server

Lenovo NVIDIA GB300 NVL72 Rack Scale AI Server
Memory960 GB LPDDR5X per tray; 17.28 TB per rack; HBM3e
Accelerators72 × NVIDIA Blackwell Ultra B300 GPUs per rack
ChipsetN/A
Processor36 × NVIDIA Grace processors per rack

NVIDIA GB300 NVL72 by HPE

NVIDIA GB300 NVL72 by HPE
Memory18.432 TB installed / 17.28 TB usable CPU memory Up to 20 TB HBM3E with 569 TB/s bandwidth
Accelerators72× NVIDIA B300 GPUs
ChipsetASPEED 2600 management controller, based on NVIDIA OpenBMC.
Processor36× NVIDIA Grace CPUs per rack

NVIDIA Vera Rubin NVL4 Module with 4 Rubin GPUs and 2 Vera CPUs

NVIDIA Vera Rubin NVL4 Module with 4 Rubin GPUs and 2 Vera CPUs
MemoryUp to 4.1 TB fast-access memory: 1.1 TB HBM4 and up to 3 TB LPDDR5X
Accelerators4× NVIDIA Rubin GPUs; sixth-generation NVLink via second-generation NVLink bridge
Chipset4× Rubin GPUs and 2× Vera CPUs connected over NVLink-C2C
Processor2 × NVIDIA Vera 88-core CPUs

NVIDIA DGX SuperPOD with DGX Vera Rubin NVL72 and Quantum-X800 InfiniBand

NVIDIA DGX SuperPOD with DGX Vera Rubin NVL72 and Quantum-X800 InfiniBand
Memory75 TB fast memory per DGX rack
Accelerators72× Rubin GPUs per rack; 8 racks per scalable unit
ChipsetN/A
Processor36× Vera CPUs per DGX rack

NVIDIA DGX SuperPOD with DGX Vera Rubin NVL72 - Spectrum-X Ethernet

NVIDIA DGX SuperPOD with DGX Vera Rubin NVL72 - Spectrum-X Ethernet
Memory75 TB fast memory per DGX rack
Accelerators72× Rubin GPUs per rack; 3 racks per scalable unit
ChipsetN/A
Processor36 × NVIDIA Vera CPUs per rack

NVIDIA DGX Vera Rubin NVL72 Rack-Scale AI System

NVIDIA DGX Vera Rubin NVL72 Rack-Scale AI System
Memory75 TB total fast memory: 20.7 TB HBM4 GPU memory and 54 TB LPDDR5X CPU memory
Accelerators72× NVIDIA Rubin GPUs; 9× NVLink switch trays
ChipsetCompute tray with 2× NVIDIA Vera Rubin Superchips (2 Rubin GPUs + 1 Vera CPU each)
Processor36 × NVIDIA Vera 88-core CPUs

NVIDIA Vera Rubin NVL72 OEM-Integrated Rack-Scale AI Platform

NVIDIA Vera Rubin NVL72 OEM-Integrated Rack-Scale AI Platform
Memory75 TB total fast-access memory: 20.7 TB HBM4 and 54 TB LPDDR5X
Accelerators72× NVIDIA Rubin GPUs; sixth-generation NVLink
ChipsetNVIDIA Vera Rubin Superchips (2 Rubin GPUs + 1 Vera CPU each)
Processor36 × NVIDIA Vera 88-core CPUs

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

NVIDIA HGX Rubin NVL8 AI Platform with 8 Rubin GPUs - Vera or x86 CPU Configuration
Memory2.3 TB HBM4 GPU memory; Vera configuration adds 1.5 TB LPDDR5X CPU memory; x86 memory is OEM-defined
Accelerators8× NVIDIA Rubin SXM GPUs; sixth-generation NVLink
ChipsetNVIDIA HGX Rubin NVL8 8-GPU SXM baseboard
Processor1 × NVIDIA Vera 88-core CPU or OEM-selected x86 CPU

NVIDIA DGX Rubin NVL8

NVIDIA DGX Rubin NVL8
Memory2.3 TB HBM4 GPU memory; system RAM: Information Not Available
Accelerators8× NVIDIA Rubin GPUs; sixth-generation NVLink
ChipsetN/A
Processor2× Intel Xeon 6776P

NVIDIA DGX SuperPOD with DGX Rubin NVL8 Systems - 72-System Scalable Unit

NVIDIA DGX SuperPOD with DGX Rubin NVL8 Systems - 72-System Scalable Unit
Memory2.3 TB GPU memory per DGX system
Accelerators8× Rubin GPUs per DGX system; 72 systems per scalable unit
ChipsetN/A
Processor2× Intel Xeon 6776P per DGX system
Dell

Dell

Reliable, high-density foundations for mission-critical AI training and inference.


Custom AI Architecture
Optimized TCO
Full-Stack Lifecycle

AI Infrastructure Solutions for Real Workloads

Leverage Uvation’s vertically integrated AI infrastructure to map validated configurations directly to your real-world power and deployment constraints.

https://cdn.uvation.com/uvationmarketplace/homepage_v2_assets/home/rewards-integration.svg

Vertically Integrated Infrastructure

Scale model training and high-throughput inference on architectures engineered for maximum GPU density and seamless deployment into AI factories or existing data centers.

https://cdn.uvation.com/uvationmarketplace/homepage_v2_assets/home/rewards-flexibility.svg

Flexible Procurement

Match commercial terms to your deployment roadmap with purchase, lease, or rental options tailored to traditional, modular, or autonomous infrastructure sites.

https://cdn.uvation.com/uvationmarketplace/homepage_v2_assets/home/rewards-extensible.svg

Sustainable Operations

Standardize lifecycle management across regions using validated modular designs that allow you to upgrade to next-generation GPUs without rebuilding your infrastructure from scratch.

Specialized Support for AI Infrastructure

From architectural guidance to complex problem solving, our experts ensure your AI environment remains optimized and resilient.

https://cdn.uvation.com/uvationmarketplace/homepage_v2_assets/home/support-sales.svg

Sourcing and Sales

Access our specialized supply chain for mission-critical GPU components and infrastructure hardware precisely when your scaling demands it.

https://cdn.uvation.com/uvationmarketplace/homepage_v2_assets/home/support-lifecycle.svg

Product Lifecycle

Protect your AI value chain with expert sourcing and proactive management, ensuring hardware continuity through every stage of the technology lifecycle.

https://cdn.uvation.com/uvationmarketplace/homepage_v2_assets/home/support-ordering.svg

Self Service Ordering

Scalable, self-service procure-ment through our marketplace, allowing you to configure and deploy specialized compute resources anytime, anywhere.

https://cdn.uvation.com/uvationmarketplace/homepage_v2_assets/home/support-financing.svg

Financing & Leasing

Access flexible capital solutions , including lease and net-term options designed to align with your specific AI deployment and growth objectives.

Engage Your Accountable Operator

Connect with our specialized team to move from validated configurations to operational AI capacity with fewer unknowns.