Uvation MarketplaceMarketplace

Lenovo ThinkSystem
SR680a V4 Server

Accelerate demanding AI workloads with eight NVIDIA B300 GPUs, 2.3 TB of HBM3e memory and up to 800 Gb/s networking. Enterprise reliability, scalable storage and Lenovo ThinkShield security make it ideal for modern AI infrastructure.

8x
NVIDIA HGX B300 NVL8 GPUs
Up to 4 TB DDR5
Memory
Dual Intel Xeon 6700P
Processors with up to 86 cores each

Rack-Scale Compute for
Demanding AI Workloads

An integrated rack architecture brings GPU compute, processor resources,
high-speed networking, and dedicated power infrastructure into one system.

High-Density GPU Architecture

Houses 72 NVIDIA Blackwell Ultra B300 GPUs across 18 compute trays.

Integrated Memory Design

Provides 960 GB LPDDR5X memory per tray, totaling 17.28 TB per rack, alongside HBM3e GPU memory.

Built-In Network Infrastructure

Incorporates ConnectX-8 network boards, BlueField-3, and NVLink Switch trays within the rack-scale configuration.

Eight-GPU NVIDIA B300 Configuration

The XD690 integrates eight NVIDIA B300 SXM GPUs within a single 10U server platform, supported by dual Intel Xeon 6 processors and high-capacity DDR5 memory.

Eight-GPU NVIDIA B300 server
8× NVIDIA B300
SXM GPUs
288 GB
memory per GPU
Up to 3 TB DDR5
memory
8× E1.S EDSFF NVMe
SSDs

Workloads

A rack-scale AI platform combining 72 NVIDIA B300 GPUs, 36 Grace CPUs, 800 Gb/s networking, and liquid cooling in a single NVLink domain.

AI Model Development

648 petaFLOPS of sparse FP8 compute and 20.7 TB of HBM3e help teams train larger models in less time.

GPU-Accelerated Computing

1,296 petaFLOPS of sparse FP4 compute speed up agentic reasoning and long-context responses.

AI Workload Execution

36 Grace CPUs with 2,592 Arm cores and 17.3 TB of LPDDR5X feed simulation, genomics, research and analytics alongside AI.

Engineered to dominate

Designed for Dense AI Infrastructure

A 10U platform combining GPU acceleration, server processing, high-speed networking, and integrated power hardware.

Exploded view of the Lenovo ThinkSystem SR680a V4 server
8× ConnectX
8 network adapters
12× 3,200 W
hot-plug PSUs
Silicon Root
of Trust and Secure Boot
Canonical Ubuntu 24.04 LTS
Server
Exploded view of the Lenovo ThinkSystem SR680a V4 server
8× ConnectX
8 network adapters
12× 3,200 W
hot-plug PSUs
Silicon Root
of Trust and Secure Boot
Canonical Ubuntu 24.04 LTS
Server

The rack, itemized

SUBSYSTEM

SPECIFICATION / CONFIGURATION DETAILS

Processor

2× Intel Xeon 6700P; up to 86 cores each

Operating System

RHEL 9.6; Ubuntu 24.04 LTS

Chipset

N/A

Accelerators / GPU

8 x NVIDIA HGX B300 NVL8 GPUs

Memory

32 DDR5 DIMMs; up to 4 TB

Primary Storage

8x 2.5-inch hot-swap drive bays

Secondary Storage

2 × M.2 NVMe SSDs

Security

TPM 2.0; Secure Boot; firmware resiliency; XClarity security

Chassis

8U chassis; 21 fans

Power Supply

6 or 8× 3,200 W; or 6 or 8× 3,800 W PSUs

Ports

8× OSFP; four USB 3.0; XCC RJ45; front video

Slots / PCIe Architecture

4× PCIe 5.0 x16 FHHL; one OCP slot

Bezel

N/A

Form Factor

8U Rack Server

Network Card

8× ConnectX-8; configurable OCP NIC

Warranty

Three-year CRU and onsite; 9×5 next-business-day response

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.