Lenovo 5WS1B61706 ThinkStation 3-Yr Premier Onsite Warranty

• Extend support coverage for enterprise AI development workstations
• Protect longer-term local LLM development environments
• Support AI proof-of-concept infrastructure lifecycle planning
• Maintain service coverage for machine-learning development nodes
• Extend warranty protection across corporate AI labs
• Support universities deploying AI research workstations
• Protect compact AI infrastructure in innovation teams
• Align workstation support with multi-year IT budgets
• Extend service planning for local generative AI platforms
• Support distributed AI development and experimentation
• Protect PGX deployments used for model development
• Improve warranty continuity across multi-year projects

• 3-Year Premier Onsite Support coverage for PGX
• Upgrades the original 1-Year Premier Onsite term
• Supports longer AI workstation deployment cycles
• Helps improve service continuity for business-critical systems
• Aligns support duration with extended asset planning
• Reduces premature replacement pressure from support expiry
• Complements PGX local AI development infrastructure
• Supports predictable enterprise IT lifecycle planning
• Applies to the Lenovo ThinkStation PGX platform
• Designed for organizations requiring extended service cover
• Supports procurement planning beyond the initial warranty term
• Helps maintain operational support across longer deployments

• Brand: Lenovo
• Product Family: ThinkStation
• Product Series: ThinkStation PGX
• Product Type: Compact AI Development Workstation
• Warranty SKU: 5WS1B61706
• Warranty Service: Premier Onsite Support
• Original Support: 1-Year Premier Onsite Support
• Upgraded Support: 3-Year Premier Onsite Support
• Warranty Upgrade: 1 Year - 3 Years
• Processor: NVIDIA Grace 20-Core Arm CPU
• CPU Architecture: NVIDIA Grace
• CPU Core Configuration: 10 Cortex-X925 + 10 Cortex-A725 cores
• GPU Architecture: NVIDIA Blackwell
• AI Performance: Up to 1,000 TOPS
• FP4 Performance: Up to 1 PFLOP with sparsity
• Unified Memory: 128GB LPDDR5x
• Memory Bus: 256-bit
• Memory Bandwidth: 273GB/s
• Storage: 1TB or 4TB NVMe M.2
• Storage Security: Self-Encrypting Drive support
• Network Interface: 10Gb Ethernet RJ-45
• High-Speed Networking: NVIDIA ConnectX-7
• High-Speed Ports: 2 × QSFP ports
• Wireless: Wi-Fi 7
• Bluetooth: Bluetooth 5.3 LE
• USB: 3 × USB4 20Gb/s
• Display Interface: HDMI 2.1a
• USB-C Display: DisplayPort 2.1 support via USB4
• Power Supply: 240W USB-C PD 3.1
• Operating System: NVIDIA DGX OS
• Alternative OS: Ubuntu Linux Pro with NVIDIA Base OS
• Security: TPM 2.0
• Secure Boot: UEFI Secure Boot
• Firmware Protection: NVIDIA Firmware Recovery
• Platform Security: NVLink-C2C Enclave
• Form Factor: Small Form Factor
• Chassis Volume: 1.13L
• Dimensions: 150 × 150 × 50.50 mm
• Starting Weight: 1.2 kg / 2.65 lb
• Multi-Node Scaling: Up to 2 PGX nodes via ConnectX-7
• Compatibility: Lenovo ThinkStation PGX
• Primary Workload: AI Development and Deployment
• Target Environment: Enterprise AI, Research, Development
• Sustainability: RoHS Compliant

WHY CHOOSE THIS PRODUCT?

• Extends service protection Moves PGX coverage from 1 to 3 years, helping align support with longer enterprise asset cycles.
• Protects AI infrastructure investment Extends Premier Onsite Support around a workstation designed for demanding local AI development.
• Improves lifecycle predictability A longer support term makes warranty planning easier for procurement, IT operations and finance teams.
• Supports local AI deployment PGX combines Grace Blackwell compute, 128GB unified memory and high-speed networking for AI-focused environments.
• Enables compact AI infrastructure The 1.13L chassis provides substantial AI compute capability without requiring a conventional large workstation footprint.
• Strengthens data protection PGX supports self-encrypting NVMe storage, TPM 2.0 and UEFI Secure Boot for enterprise-oriented platform security.
• Supports future scaling Two PGX systems can be connected through integrated NVIDIA ConnectX-7 technology for larger model workloads.