MSI XpertStation WS300 NVIDIA DGX Station AI Supercomputer

MSI XpertStation WS300 NVIDIA DGX Station with Grace Blackwell Ultra, 748GB Coherent Memory, 1.92TB RAID 1 & Dual 400G ConnectX-8 Networking -  Local Warranty 

NVIDIA DGX Station Performance for Enterprise AI Innovation

The MSI XpertStation WS300 is an enterprise-class AI supercomputer built on the NVIDIA DGX Station platform, delivering breakthrough performance for artificial intelligence, machine learning, and high-performance computing. Powered by the NVIDIA Grace Blackwell Ultra (GB300) Superchip, the system provides up to 20 PFLOPS of AI compute, enabling organisations to train, fine-tune, and deploy large language models (LLMs), AI agents, generative AI applications, computer vision models, and scientific simulations directly within their own secure infrastructure. It is purpose-built for enterprises, research institutions, universities, healthcare organisations, and AI development teams seeking workstation-class performance without relying on cloud resources.

748GB Coherent Unified Memory for Large AI Models

Equipped with an exceptional 748GB of coherent unified memory, the MSI XpertStation WS300 enables AI models and datasets to be processed efficiently without the memory limitations of conventional GPU workstations. This advanced memory architecture accelerates large-scale inferencing, retrieval-augmented generation (RAG), AI agent frameworks, engineering simulations, financial modelling, digital twins, and complex scientific research, delivering faster insights and improved productivity.

Enterprise Storage and Ultra-High-Speed 400G Networking

The workstation includes 1.92TB enterprise NVMe storage configured in RAID 1 for enhanced data protection and system reliability. Dual NVIDIA ConnectX®-8 400GbE networking provides ultra-low latency communication and exceptional bandwidth, allowing seamless integration into AI clusters and distributed computing environments. Whether deployed as a standalone AI workstation or connected to larger AI infrastructures, the XpertStation WS300 delivers the networking performance required for today's most demanding workloads.

Built for Continuous Enterprise AI Operations

Designed for mission-critical environments, the MSI XpertStation WS300 incorporates enterprise-grade reliability, advanced thermal management, remote administration capabilities, and NVIDIA Multi-Instance GPU (MIG) support for secure multi-user workloads. It is an ideal platform for private AI clouds, AI model development, cybersecurity research, healthcare analytics, engineering design, autonomous systems, and large-scale enterprise AI deployments where performance, scalability, and data sovereignty are essential.

Technical Specifications

General

Model: MSI XpertStation WS300
Product Name: MSI XpertStation WS300 NVIDIA DGX Station with Grace Blackwell Ultra, 748GB Coherent Memory, 1.92TB RAID 1 & Dual 400G ConnectX-8 Networking
Manufacturer: MSI
Product Series: MSI XpertStation
Product Type: Enterprise AI Workstation / NVIDIA DGX Station-Class AI System
Compute Platform: NVIDIA Grace Blackwell Ultra
Processor Architecture: NVIDIA Grace CPU + NVIDIA Blackwell Ultra GPU
CPU Platform: Single NVIDIA Grace CPU Superchip
GPU Platform: Single NVIDIA Blackwell Ultra GPU
System Architecture: Unified CPU-GPU Accelerated Computing Platform
Memory Architecture: NVIDIA Coherent Unified Memory
Total Coherent Memory: 748GB
Operating System: Ubuntu 24.04 LTS
Pre-Installed Software: NVIDIA AI Developer Tools
Primary Storage: 1.92TB NVMe RAID 1
Storage Expansion: 2 × PCIe 6.0 x4 M.2 2280 Slots Available
High-Speed Networking: Dual NVIDIA ConnectX-8 400G QSFP112
Dedicated Management: 1000Base-T Server Management Port
Remote Management Controller: ASPEED AST2600
Management Protocols: IPMI 2.0 and DMTF Redfish
Security: Hardware Root of Trust and TPM 2.0
Target Deployment: Enterprise AI Development, AI Research, Model Training, Model Fine-Tuning, Inferencing, Scientific Computing and High-Performance Computing
Target Users: Enterprises, Government Agencies, Universities, Research Institutions, Data Scientists, AI Developers and Engineering Teams
Warranty: MSI Enterprise Warranty, Configuration Dependent

Processor Specifications

Processor Platform: NVIDIA Grace CPU Superchip
Processor Quantity: 1
CPU Architecture: Arm
CPU Core Architecture: Arm Neoverse V2
CPU Core Count: 72 Cores
CPU Type: NVIDIA Grace
System-on-Chip Architecture: Supported
High-Bandwidth CPU-GPU Interconnect: Supported
Unified CPU-GPU Address Space: Supported
Coherent CPU-GPU Memory Access: Supported
Enterprise Parallel Computing: Supported
Large-Scale AI Processing: Supported
High-Performance Computing: Supported
Scientific Computing: Supported
Multi-Threaded Workloads: Supported
Continuous Enterprise Workloads: Supported
Optimized for NVIDIA AI Software Stack
Optimized for Generative AI Development
Optimized for Large Language Model Processing

AI Performance

AI Compute Architecture: NVIDIA Blackwell Ultra
AI Accelerator Quantity: 1
GPU Type: NVIDIA Blackwell Ultra GPU
Transformer Engine: NVIDIA Blackwell Ultra Transformer Engine
Tensor Core Acceleration: Supported
Mixed-Precision AI Computing: Supported
FP4 AI Computing: Supported
FP8 AI Computing: Supported
FP16 AI Computing: Supported
BF16 AI Computing: Supported
INT8 Inferencing: Supported
Dynamic Precision Management: Supported
Large Language Model Acceleration: Supported
Foundation Model Training: Supported
Foundation Model Fine-Tuning: Supported
Generative AI Acceleration: Supported
AI Agent Development: Supported
Reasoning Model Processing: Supported
Multimodal AI Processing: Supported
Retrieval-Augmented Generation: Supported
Production AI Inferencing: Supported
Scientific AI Processing: Supported
Distributed AI Workloads: Supported
Enterprise AI Development: Supported

AI Features

Large Language Model Training
Large Language Model Fine-Tuning
Foundation Model Development
Generative AI Applications
AI Agent Development
Agentic AI Workloads
Reasoning Models
Retrieval-Augmented Generation
Multimodal AI Processing
Natural Language Processing
Computer Vision
Speech Recognition
Speech Generation
Recommendation Systems
Predictive Analytics
Digital Twin Simulation
Autonomous Systems Development
Robotics AI
Scientific Machine Learning
Healthcare AI Research
Drug Discovery
Molecular Modelling
Financial Modelling
Cybersecurity AI
Manufacturing AI
Enterprise Knowledge Assistants
Private AI Deployment
On-Premise AI Development
Cloud-Native AI Deployment
Production AI Inferencing
Multi-User AI Development
NVIDIA AI Developer Environment

Memory Specifications

Total System Memory: 748GB
Memory Architecture: NVIDIA Coherent Unified Memory
CPU-GPU Shared Memory Pool: Supported
Unified Address Space: Supported
Memory Coherency: Supported
Direct CPU and GPU Memory Access: Supported
High-Bandwidth Memory Architecture: Supported
Low-Latency CPU-GPU Data Access: Supported
Reduced Memory Copy Operations: Supported
Large AI Model Loading: Supported
Large Context Window Processing: Supported
Multi-Billion Parameter Model Support: Supported
Large Dataset Processing: Supported
AI Model Fine-Tuning Support: Supported
Enterprise Inferencing Support: Supported
Retrieval-Augmented Generation Support: Supported
Vector Database Processing: Supported
Multi-Tenant Memory Allocation: Supported
Continuous AI Pipeline Processing: Supported
Scientific Simulation Support: Supported
High-Performance Data Analytics: Supported

Storage Specifications

Primary Storage Technology: NVMe Solid-State Drive
Primary Storage Interface: PCI Express 5.0 x4
Primary Storage Form Factor: M.2 2280
Installed PCIe 5.0 M.2 Slots: 2
Installed Storage Configuration: 2 × M.2 2280 NVMe SSDs
Installed Usable RAID Capacity: 1.92TB
RAID Configuration: RAID 1
RAID Type: Mirrored Storage
Storage Redundancy: Supported Through RAID 1
Data Protection: Drive Mirroring
Primary Storage Slot Status: Fully Populated
PCIe 5.0 M.2 Slot Capacity: 2 × M.2 2280 NVMe PCIe 5.0 x4
PCIe 6.0 M.2 Slot Capacity: 2 × M.2 2280 NVMe PCIe 6.0 x4
PCIe 6.0 M.2 Slot Status: Open and Available for Expansion
Total M.2 Storage Slots: 4
Boot Drive Support: Supported
Enterprise NVMe SSD Support: Supported
High-Speed AI Dataset Loading: Supported
Large AI Model Repository Support: Supported
AI Model Checkpoint Storage: Supported
Vector Database Storage: Supported
Training Dataset Storage: Supported
Container Image Storage: Supported
High-Speed Data Preprocessing: Supported
Continuous Read and Write Workloads: Supported
Storage Expansion: Supported
Storage Health Monitoring: Supported
Remote Storage Status Monitoring: Supported
Future PCIe 6.0 Storage Expansion: Supported

RAID Features

RAID Level: RAID 1
RAID Configuration: Two-Drive Mirrored Array
Installed RAID Capacity: 1.92TB
Primary Purpose: Data Redundancy and Operating Continuity
Drive Failure Protection: Supported for a Single Drive Failure
Mirrored Data Storage: Supported
Enterprise Boot Volume Protection: Supported
AI Development Environment Protection: Supported
Model and Configuration Data Protection: Supported
Storage Rebuild Support: Configuration Dependent
RAID Monitoring: Supported
Drive Health Monitoring: Supported
RAID Status Reporting: Supported

Graphics Specifications

GPU Platform: NVIDIA Blackwell Ultra
GPU Quantity: 1
GPU Type: Enterprise AI and HPC Accelerator
Tensor Cores: NVIDIA Tensor Cores
Transformer Engine: Supported
CUDA Parallel Computing: Supported
CUDA Toolkit Compatibility: Supported
NVIDIA cuDNN Compatibility: Supported
NVIDIA TensorRT Compatibility: Supported
NVIDIA RAPIDS Compatibility: Supported
NVIDIA NeMo Compatibility: Supported
NVIDIA NGC Container Compatibility: Supported
NVIDIA AI Enterprise Compatibility: Supported
GPUDirect RDMA: Supported
GPUDirect Storage: Supported
Mixed-Precision Computing: Supported
Large Language Model Acceleration: Supported
Generative AI Acceleration: Supported
Scientific Simulation Acceleration: Supported
High-Performance Data Analytics: Supported
AI Training: Supported
AI Fine-Tuning: Supported
Production Inferencing: Supported
Multi-Modal AI Processing: Supported

Networking Specifications

High-Speed Network Controller: NVIDIA ConnectX-8 SuperNIC
High-Speed Network Controller Quantity: 2
High-Speed Network Ports: 2 × 400G QSFP112
Connector Type: QSFP112
Maximum Per-Port Bandwidth: Up to 400Gbps
Maximum Aggregate Network Bandwidth: Up to 800Gbps
Ethernet Support: Supported
RDMA Support: Supported
RoCE Support: Supported
GPUDirect RDMA: Supported
GPUDirect Storage: Supported
Ultra-Low-Latency Networking: Supported
Distributed AI Training: Supported
AI Cluster Interconnect: Supported
High-Speed Storage Fabric: Supported
Scale-Out AI Infrastructure: Supported
Enterprise AI Fabric Integration: Supported
Multi-Node Model Training: Supported
High-Performance Computing Cluster Support: Supported
Low-Latency Data Transfer: Supported
High-Throughput Model Distribution: Supported

Dedicated Management Networking

Dedicated Management Port: 1 × 1000Base-T
Management Network Speed: 1GbE
Connector Type: RJ-45
Management Port Function: Dedicated Out-of-Band Server Management
Management Network Isolation: Supported
Remote Management Without Host Operating System: Supported
Remote Power Control: Supported
Remote Console Access: Supported
Remote Firmware Management: Supported
System Health Monitoring: Supported
Event Logging: Supported
Enterprise Datacentre Integration: Supported

Ports & Expansion

2 × 400G QSFP112 Ports with NVIDIA ConnectX-8 SuperNICs
1 × 1000Base-T Dedicated Server Management Port
2 × M.2 2280 PCIe 5.0 x4 NVMe Slots, Populated
2 × M.2 2280 PCIe 6.0 x4 NVMe Slots, Available for Expansion
3 × PCIe 5.0 Expansion Slots
Enterprise Network Expansion Support
High-Speed Storage Expansion Support
Additional Accelerator Expansion: Configuration Dependent
Future PCIe Device Expansion: Supported
Dedicated BMC Management Interface: Supported
Datacentre Network Integration: Supported
AI Cluster Expansion: Supported
High-Performance Storage Connectivity: Supported

PCIe Expansion Specifications

PCIe Expansion Slot Count: 3
PCIe Generation: PCI Express 5.0
Expansion Use Cases: High-Speed Networking, Storage Controllers, Data Acquisition, Additional Accelerators and Enterprise I/O
PCIe 5.0 Device Support: Supported
Enterprise Add-In Card Support: Supported
High-Bandwidth Peripheral Support: Supported
Low-Latency Expansion: Supported
Future Hardware Expansion: Supported
GPU and Accelerator Expansion: Configuration Dependent
Storage Controller Expansion: Supported
Network Interface Expansion: Supported
Specialized AI Hardware Expansion: Supported

Operating System

Pre-Installed Operating System: Ubuntu 24.04 LTS
Operating System Type: 64-Bit Linux
Long-Term Support Release: Supported
Enterprise Linux Environment: Supported
NVIDIA Driver Stack: Pre-Installed
NVIDIA AI Developer Tools: Pre-Installed
CUDA Development Environment: Pre-Installed or Ready
NVIDIA Container Toolkit: Supported
Docker Support: Supported
Kubernetes Support: Supported
Python AI Development: Supported
AI Framework Compatibility: Supported
Remote Administration: Supported
Command-Line Administration: Supported
Enterprise Package Management: Supported
Security Update Support: Supported
Long-Term Software Maintenance: Supported

Pre-Installed NVIDIA AI Developer Tools

NVIDIA GPU Drivers
NVIDIA CUDA Toolkit
NVIDIA Container Toolkit
NVIDIA NGC Container Support
NVIDIA TensorRT
NVIDIA cuDNN
NVIDIA Triton Inference Server Support
NVIDIA NeMo Support
NVIDIA RAPIDS Support
PyTorch Support
TensorFlow Support
JAX Support
ONNX Runtime Support
Hugging Face Transformers Support
Docker Container Support
Kubernetes Integration Support
AI Model Development Tools
AI Model Optimization Tools
AI Model Serving Tools
Enterprise AI Deployment Tools

Management Controller

Management Controller: ASPEED AST2600
Management Type: Integrated Baseboard Management Controller
Out-of-Band Management: Supported
IPMI Version: IPMI 2.0
DMTF Redfish Support: Supported
Remote KVM Access: Supported
Remote Console Access: Supported
Remote Power On: Supported
Remote Power Off: Supported
Remote Power Cycle: Supported
Remote System Reset: Supported
Remote Firmware Updates: Supported
Remote BIOS Configuration: Supported
Virtual Media Support: Configuration Dependent
System Event Log: Supported
Sensor Monitoring: Supported
Processor Monitoring: Supported
GPU Monitoring: Supported
Memory Monitoring: Supported
Storage Monitoring: Supported
Network Monitoring: Supported
Temperature Monitoring: Supported
Fan Monitoring: Supported
Voltage Monitoring: Supported
Power Monitoring: Supported
Hardware Inventory: Supported
Enterprise Fleet Management: Supported
Management API Integration: Supported
Automated Datacentre Management: Supported

IPMI 2.0 Features

Remote Hardware Monitoring
Out-of-Band System Management
Remote Power Control
System Event Logging
Sensor Data Monitoring
Hardware Alerting
Independent Management Processor
Management Without Host OS Access
Remote Troubleshooting
Enterprise Management Platform Integration
User and Role Management
Secure Management Sessions
Hardware Status Reporting
Automated Alert Generation

DMTF Redfish Features

RESTful Management API
Standards-Based Server Management
Remote Hardware Inventory
Remote Power Management
Remote Firmware Management
System Health Reporting
Storage Monitoring
Network Monitoring
Thermal Monitoring
Power Consumption Monitoring
Automation and Orchestration Support
Datacentre Management Software Integration
Enterprise Fleet Management Integration
Scripted Management Operations
Secure API-Based Administration

Security Features

Hardware Root of Trust
Trusted Platform Module 2.0
Secure Boot Support
Firmware Integrity Protection
Hardware-Based Platform Validation
Secure Firmware Update Support
BIOS Security Support
BMC User Authentication
Role-Based Access Control
Secure Remote Management
Encrypted Management Sessions
Management Network Isolation
System Event Logging
Audit Logging
Secure Boot Chain
Platform Integrity Verification
Enterprise Authentication Integration
Data-at-Rest Protection: Drive Dependent
RAID 1 Data Redundancy
Secure Operating System Environment
Ubuntu Security Updates
NVIDIA Driver Security Updates
Physical Security Features: Configuration Dependent

Hardware Root of Trust

Hardware-Based Boot Validation
Firmware Authenticity Verification
Platform Integrity Checking
Protection Against Unauthorized Firmware
Secure Firmware Recovery Support
Trusted Startup Process
BMC Firmware Security
BIOS Integrity Protection
Enterprise Platform Attestation
Secure Component Authentication
Protection Against Low-Level System Tampering

Trusted Platform Module

TPM Version: TPM 2.0
Hardware-Backed Cryptographic Storage
Platform Identity Protection
Secure Key Storage
Measured Boot Support
Device Authentication Support
Disk Encryption Integration
Operating System Security Integration
Enterprise Certificate Support
Secure Credential Protection
Platform Integrity Measurement

Virtualisation & Container Support

Docker Support
Kubernetes Support
NVIDIA Container Toolkit
NVIDIA NGC Containers
Containerised AI Development
Containerised Model Training
Containerised Inferencing
Virtual Machine Support: Configuration Dependent
Cloud-Native AI Deployment
Enterprise MLOps Platforms
Workload Orchestration
Multi-User Development Environments
Resource Scheduling
AI Development Sandboxes
DevOps and MLOps Integration
Private Cloud Integration
Hybrid Cloud Integration

Software & AI Framework Support

NVIDIA CUDA Toolkit
NVIDIA cuDNN
NVIDIA TensorRT
NVIDIA Triton Inference Server
NVIDIA NeMo
NVIDIA RAPIDS
NVIDIA NGC Containers
PyTorch
TensorFlow
JAX
ONNX Runtime
Hugging Face Transformers
DeepSpeed
Megatron-LM
Docker
Kubernetes
MLflow
Ray
Apache Spark
Vector Database Platforms
Enterprise MLOps Platforms
AI Model Serving Frameworks
Distributed Training Frameworks
Scientific Computing Libraries
Python Development Tools
C++ CUDA Development Tools

Cooling & Reliability

Enterprise Workstation Cooling Architecture
High-Efficiency Thermal Management
CPU Thermal Monitoring
GPU Thermal Monitoring
Memory Thermal Monitoring
Storage Thermal Monitoring
Variable-Speed Fan Control
Automated Fan Management
Thermal Protection
Hardware Health Alerts
Continuous AI Workload Support
Long-Duration AI Training Support
Enterprise-Grade Components
24×7 Operational Design
High-Bandwidth Component Cooling
PCIe Expansion Cooling Support
High-Speed NVMe Cooling Support
Datacentre and Laboratory Deployment Ready

Power Specifications

Power Supply Type: Enterprise High-Efficiency Power Supply
Power Supply Capacity: Configuration Dependent
Input Voltage: Configuration Dependent
Power Consumption: Workload and Configuration Dependent
Remote Power Monitoring: Supported
Remote Power Control: Supported
Power Usage Reporting: Supported
Power Fault Monitoring: Supported
Continuous High-Performance Operation: Supported
High-Load GPU Workload Support: Supported
Enterprise Power Management: Supported
Datacentre Power Monitoring Integration: Supported

Compatibility

Compatible AI Workloads:

Large Language Models
Foundation Models
Generative AI
AI Agents
Agentic AI
Retrieval-Augmented Generation
Reasoning Models
Multimodal AI
Computer Vision
Natural Language Processing
Speech AI
Recommendation Engines
Scientific Machine Learning
Predictive Analytics
Digital Twins
Autonomous Systems
Enterprise AI Inferencing
Model Training
Model Fine-Tuning
Model Optimization
Model Serving

Compatible Development Frameworks:

PyTorch
TensorFlow
JAX
ONNX Runtime
Hugging Face Transformers
DeepSpeed
NVIDIA NeMo
NVIDIA TensorRT
NVIDIA Triton Inference Server
NVIDIA RAPIDS

Compatible Deployment Platforms:

Ubuntu 24.04 LTS
Docker
Kubernetes
NVIDIA NGC
Enterprise MLOps Platforms
Private Cloud Infrastructure
Hybrid Cloud Infrastructure
AI Factory Infrastructure
Scientific Computing Environments
Enterprise DevOps Environments

Compatible Network Environments:

400G Ethernet Fabrics
NVIDIA ConnectX-8 Networking
QSFP112 Infrastructure
RDMA Networks
RoCE Networks
High-Performance Storage Networks
Distributed AI Clusters
High-Performance Computing Clusters

Design & Construction

Enterprise AI Workstation Design
Professional NVIDIA DGX Station-Class Architecture
High-Density Compute Platform
High-Bandwidth Internal Architecture
Serviceable Internal Components
Enterprise Cable Management
Dedicated Management Controller
High-Speed Expansion Architecture
Continuous 24×7 Operational Design
Remote Serviceability
Enterprise Laboratory Deployment Ready
Datacentre Integration Ready
AI Research Environment Ready
Office and Engineering Workspace Deployment Ready

Environmental Specifications

Operating Environment: Enterprise Office, Research Laboratory or Datacentre
Operating Temperature: Configuration Dependent
Storage Temperature: Configuration Dependent
Operating Humidity: Configuration Dependent
Storage Humidity: Configuration Dependent
Maximum Operating Altitude: Configuration Dependent
Thermal Monitoring: Supported
Environmental Monitoring: Supported
Continuous Operation: Supported
Compliance and Certifications: Region and Configuration Dependent

Physical Specifications

Product Type: Enterprise AI Workstation
Form Factor: Professional AI Station
Chassis Colour: Configuration Dependent
Dimensions: Configuration Dependent
System Weight: Configuration Dependent
Service Access: Enterprise Service Design
Expansion Access: Supported
Internal M.2 Access: Supported
PCIe Expansion Access: Supported
Management Port Access: Dedicated Rear I/O
High-Speed Network Port Access: Dual QSFP112

Package Contents

MSI XpertStation WS300 AI System
NVIDIA Grace CPU Superchip
NVIDIA Blackwell Ultra GPU
748GB Coherent Unified Memory
1.92TB RAID 1 NVMe Storage Configuration
Ubuntu 24.04 LTS
NVIDIA AI Developer Tools
Power Cable
Quick Installation Guide
Safety Documentation
Warranty Documentation
Enterprise Support Information

Key Features

Single NVIDIA Grace CPU Superchip
72 Arm Neoverse V2 CPU Cores
Single NVIDIA Blackwell Ultra GPU
748GB Coherent Unified Memory
Unified CPU-GPU Memory Architecture
2 × M.2 2280 PCIe 5.0 x4 NVMe Slots
Installed 1.92TB RAID 1 Storage
2 × M.2 2280 PCIe 6.0 x4 NVMe Expansion Slots
3 × PCIe 5.0 Expansion Slots
Ubuntu 24.04 LTS Pre-Installed
NVIDIA AI Developer Tools Pre-Installed
2 × 400G QSFP112 Network Ports
Dual NVIDIA ConnectX-8 SuperNICs
Up to 800Gbps Aggregate High-Speed Networking
1 × Dedicated 1000Base-T Management Port
ASPEED AST2600 BMC
IPMI 2.0 Support
DMTF Redfish Support
Hardware Root of Trust
TPM 2.0 Support
Enterprise Remote Management
AI Training and Fine-Tuning Ready
Production Inferencing Ready
High-Performance Computing Ready
Enterprise AI Development Platform

Performance & Features

The MSI XpertStation WS300 is an enterprise AI workstation built around the NVIDIA Grace Blackwell Ultra platform. It combines a single NVIDIA Grace CPU Superchip featuring 72 Arm Neoverse V2 cores with a single NVIDIA Blackwell Ultra GPU, providing a tightly integrated computing architecture for large-scale AI development, model training, fine-tuning, inferencing and scientific computing.

The system includes 748GB of coherent unified memory, allowing the NVIDIA Grace CPU and Blackwell Ultra GPU to access a shared memory pool. This architecture reduces the need to copy data between separate processor and graphics memory spaces, enabling more efficient processing of large language models, long-context workloads, multimodal models, vector databases and memory-intensive scientific applications.

Primary storage is configured with two M.2 2280 NVMe PCIe 5.0 x4 drives in a 1.92TB RAID 1 array. RAID 1 mirrors data across both drives to improve storage resilience and protect the operating system, AI development tools, model configurations and application data against the failure of a single drive.

For future storage expansion, the workstation includes two additional M.2 2280 NVMe PCIe 6.0 x4 slots. These slots are open and available for high-speed storage upgrades, enabling organisations to expand capacity for AI datasets, model checkpoints, vector databases, container images and local inference repositories.

Three PCIe 5.0 expansion slots provide additional flexibility for enterprise networking adapters, storage controllers, specialist data acquisition cards and other high-bandwidth PCIe devices. This allows the workstation to be adapted for different AI, engineering, research and high-performance computing environments.

The MSI XpertStation WS300 includes dual NVIDIA ConnectX-8 SuperNICs with two 400G QSFP112 ports. The network interfaces provide up to 800Gbps of aggregate high-speed connectivity for distributed AI training, cluster communication, high-performance storage fabrics and low-latency data transfer between AI systems.

A dedicated 1000Base-T server management port connects directly to the integrated ASPEED AST2600 baseboard management controller. This provides independent out-of-band access for remote power control, hardware monitoring, system event logging, diagnostics, firmware management and remote administration even when the main operating system is unavailable.

Support for IPMI 2.0 and DMTF Redfish enables integration with enterprise datacentre management platforms, automation tools and infrastructure monitoring systems. Administrators can use standards-based interfaces to manage hardware inventory, monitor system health, automate lifecycle operations and remotely troubleshoot the workstation.

Hardware Root of Trust and TPM 2.0 provide hardware-backed platform protection, firmware validation, secure key storage, measured boot and operating system security integration. These capabilities help protect the workstation against unauthorised firmware changes and support enterprise security requirements.

Ubuntu 24.04 LTS is pre-installed together with NVIDIA AI Developer Tools, providing a ready-to-use software environment for CUDA development, AI framework deployment, containerised workloads, model optimisation and production inference. The platform supports leading AI frameworks including PyTorch, TensorFlow, JAX, Hugging Face Transformers, TensorRT, Triton Inference Server, NeMo and RAPIDS.

Typical Use Cases

Large Language Model Training
Large Language Model Fine-Tuning
Foundation Model Development
Generative AI Application Development
AI Agent Development
Agentic AI Deployment
Retrieval-Augmented Generation
Reasoning Model Processing
Multimodal AI
Production AI Inferencing
Machine Learning
Deep Learning
Computer Vision
Natural Language Processing
Speech AI
Recommendation Systems
Enterprise Knowledge Assistants
Cybersecurity AI
Financial Services AI
Healthcare AI Research
Drug Discovery
Molecular Simulation
Scientific Computing
High-Performance Computing
Digital Twin Simulation
Manufacturing AI
Autonomous Systems Development
Robotics Research
University Research Laboratories
Government AI Infrastructure
Private AI Development
On-Premise AI Deployment
Hybrid Cloud AI Development
Enterprise MLOps
Multi-User AI Development
High-Speed AI Cluster Nodes
Distributed AI Training
Enterprise AI Factories