Lenovo ThinkSystem SR675 GPU Server

Lenovo ThinkSystem SR675 GPU Server (Advance CTO, Please Enquire for Further Information) - Local Warranty

High-Performance GPU Server for AI and Accelerated Computing

Lenovo ThinkSystem SR675 GPU Server is an enterprise-grade accelerated computing platform designed for artificial intelligence, machine learning, high-performance computing (HPC), virtualization, and data-intensive workloads. Built for modern AI infrastructure and enterprise data centers, the SR675 delivers scalable GPU acceleration for advanced compute environments.

Optimized for AI, Deep Learning and HPC Workloads

The Lenovo ThinkSystem SR675 GPU Server supports demanding applications including AI model training, generative AI, large language models (LLMs), deep learning, rendering, scientific simulations, analytics, and GPU-intensive computing workloads. Its architecture enables organizations to accelerate data processing and AI development efficiently.

Flexible Configure-To-Order Infrastructure Platform

As an Advance CTO (Configure-To-Order) solution, the Lenovo ThinkSystem SR675 GPU Server supports customizable configurations including AMD EPYC processors, GPU accelerators, memory, storage, networking, and expansion options. Organizations can tailor infrastructure according to AI clusters, HPC environments, virtualization platforms, and analytics deployments.

Built for Enterprise AI and Data Center Deployment

Designed for enterprises, AI labs, research institutions, universities, and service providers, the ThinkSystem SR675 GPU Server provides scalable infrastructure for accelerated computing environments. Its rack-mounted architecture supports efficient deployment while delivering enterprise-grade reliability and operational efficiency.

Technical Specifications

General
Model: ThinkSystem SR675 V3
Product Name: Lenovo ThinkSystem SR675 GPU Server
Server Type: GPU Accelerated Rack Server
Form Factor: 3U Rack Server
Deployment: AI, Machine Learning, Deep Learning, HPC, Virtualization, Data Analytics
Configuration Type: Advanced CTO (Configure-To-Order)
Processor Platform: AMD EPYC™ 9004 Series Processors
Architecture: Accelerated Computing Platform
Target Workloads: AI Training, AI Inferencing, HPC Simulation, VDI, Scientific Research, Rendering, Generative AI

Processor Specifications
Processor Support: Up to 2 x AMD EPYC™ 9004 Series Processors
Maximum CPU Cores: Up to 128 Cores per Processor
Maximum Total Cores: Up to 256 Cores
Processor Socket: SP5
Processor Architecture: AMD EPYC Genoa / Bergamo
Maximum Processor TDP: Up to 400W per CPU

GPU Acceleration Specifications
GPU Support: Up to 8 x Double Width GPUs
GPU Architecture: High Density GPU Platform

Supported GPU Options:
• NVIDIA H100 Tensor Core GPU
• NVIDIA H200 GPU
• NVIDIA L40S GPU
• NVIDIA RTX Enterprise GPUs
• AI and HPC Accelerators

GPU Deployment Support:
• AI Model Training
• Machine Learning
• Deep Learning
• Generative AI
• Large Language Models (LLM)
• Scientific Simulation

GPU Interconnect: High Bandwidth Accelerator Architecture
Accelerated Computing Support: Supported
AI Optimization: Supported

Memory Specifications
Memory Technology: DDR5 ECC RDIMM
DIMM Slots: Up to 24 DIMM Slots
Maximum Memory Capacity: Up to 6TB
Memory Speed: Up to 4800 MT/s (processor dependent)
Memory Protection: ECC Supported

Storage Specifications
Storage Architecture: CTO Configurable
Supported Storage Types:
• NVMe SSD
• SAS SSD
• SATA SSD
• HDD

Hot Swap Drives: Supported
RAID Support: Supported
Flexible Storage Configuration: Supported

Expansion & PCIe
PCIe Generation: PCIe Gen5
Expansion Slots: Multiple PCIe Gen5 Slots
GPU Expansion: High Density Accelerator Support
Networking Expansion: OCP / PCIe Network Modules Supported

Networking
Network Interface: CTO Dependent
Supported Network Options:
• Ethernet Adapters
• High Speed AI Networking
• InfiniBand Adapters
• OCP Network Modules

Remote Management: Lenovo XClarity Controller 2 (XCC2)
Out-of-Band Management: Supported

Cooling & Thermal Design
Cooling Type: Enterprise High Performance Cooling
Thermal Architecture: GPU Optimized Airflow Design
High Density Compute Support: Supported
Redundant Cooling: Supported

Power Specifications
Power Supply Type: Redundant Hot Swap Power Supplies
Power Redundancy: Supported
Power Efficiency: 80 PLUS Platinum / Titanium (configuration dependent)

Management & Security
Management Platform: Lenovo XClarity Controller 2 (XCC2)
Remote Monitoring: Supported
Firmware Management: Supported
Server Health Monitoring: Supported
Secure Boot: Supported
TPM Security: Supported
Encryption Support: Supported

Virtualization Support
Supported Hypervisors:
• VMware vSphere
• Microsoft Hyper-V
• Linux KVM
• Red Hat Virtualization

Cloud Deployment: Supported
Container Platforms: Kubernetes Ready

Operating System Compatibility
Supported Operating Systems:
• Microsoft Windows Server
• Red Hat Enterprise Linux
• Ubuntu Server
• SUSE Linux Enterprise Server
• VMware ESXi

Design & Construction
Form Factor: 3U Rack Mount
Rack Compatibility: Standard Enterprise Rack
Deployment Environment: AI Infrastructure / Data Center / HPC Environment
Construction: Enterprise Grade Server Chassis

Performance & Features
GPU optimized architecture for AI and HPC workloads
Supports up to eight enterprise GPUs
AMD EPYC high-core-count processor platform
PCIe Gen5 high-speed expansion architecture
Large DDR5 memory support for AI datasets
Enterprise lifecycle management platform
Optimized for generative AI and LLM workloads
Scalable accelerated computing infrastructure

Package Contents
Lenovo ThinkSystem SR675 GPU Server Chassis
Power Supply Units (configuration dependent)
Rail Kit (optional)
Documentation
CTO Components according to final build

Key Features
3U GPU accelerated rack server
Supports up to 8 enterprise GPUs
AMD EPYC processor platform
DDR5 ECC memory architecture
PCIe Gen5 expansion support
AI and HPC optimized deployment
Enterprise remote management
Advanced CTO configurable platform

Typical Use Cases
AI model training
Machine learning workloads
Generative AI environments
Large Language Models (LLM)
Scientific simulations
High Performance Computing (HPC)
Virtual desktop infrastructure
Research computing environments
Enterprise AI deployment

How to Configurate Server?

Step 1:
What is the Server Form Factor? Tower or Rackmount Server?
What is the number of CPU Core or Ram or HDD usable size require?
What is the number of CPU Processor? 1S or 2S or 4S?
What is the RAID Configuration? Any specific HDD or SSD? 7.2K or 10K?
What is the specific NIC is required? Base-T 1G or SFP+? Or Fibre Channel?
What is the number of year maintenance support required? Minimum 3 years

Step 2:
How many number of unit Server?
Any VM workload? Required any VMware license? Or Window OS Hyper-V?
How do you store data currently? Any NAS or SAN Storage?
How do you doing the backup? Required any backup proposal? IDPA 8TB?
Do you require burn-in test services?
Do you need hard disk retention services?