Lenovo-ThinkSystem SR670 V2 GPU 3U Rack Server with Xeon CPU 32GB DDR4 Memory Stock & Used Products Available
Here’s a detailed comparison table for the Lenovo ThinkSystem SR670 V2 rack server, comparing its key configurations and features:
| Feature | ThinkSystem SR670 V2 (2U) | ThinkSystem SR670 V2 (4U) | Competitor Comparison |
|---|---|---|---|
| Form Factor | 2U Rack Server | 4U Rack Server | (vs HPE/Dell 2U/4U equivalents) |
| Processor Support | Dual 3rd Gen Intel Xeon Scalable | Dual 3rd Gen Intel Xeon Scalable | Similar to Dell R750/R750xa |
| Max CPU Cores | Up to 40 cores/socket | Up to 40 cores/socket | Matches HPE DL380 Gen10+ |
| Memory Capacity | 6TB DDR4 (32 DIMM slots) | 6TB DDR4 (32 DIMM slots) | Higher than most 2U competitors |
| Storage Options | 24x 2.5″ or 12x 3.5″ bays | 24x 2.5″ or 12x 3.5″ bays | More flexible than Dell R750 |
| PCIe Generation | PCIe 4.0 | PCIe 4.0 | Standard for this generation |
| PCIe Slots | 8 slots | 10 slots | More than HPE DL380 |
| GPU Support | Up to 4x GPUs (2U) | Up to 8x GPUs (4U) | 4U version beats Dell R750xa |
| Drive Bays | 12x NVMe + 12x SAS/SATA | 16x NVMe + 8x SAS/SATA | Better NVMe ratio than competitors |
| Network Options | 100GbE (standard) | 200GbE (optional) | Comparable to Cisco UCS |
| Power Supplies | 2000W (80+ Platinum) | 2400W (80+ Titanium) | More efficient than HPE options |
| Cooling System | Optimized airflow | Rear-door heat exchanger | Advanced cooling vs competitors |
| Use Cases | Virtualization, databases | AI/ML, HPC | Broader use case coverage |
| Ideal Workloads | Enterprise applications | GPU-accelerated computing | More versatile than Dell/HPE |
Lenovo ThinkSystem SR670 V2 Rack Server: Power and Flexibility Redefined
The Lenovo ThinkSystem SR670 V2 is a high-performance, GPU-ready rack server designed for AI, HPC, virtualization, and enterprise workloads. Available in 2U and 4U configurations, it delivers exceptional compute density, storage flexibility, and GPU acceleration to meet demanding business needs.
Key Features & Benefits
1. Unmatched Performance
- Dual 3rd Gen Intel Xeon Scalable Processors (up to 40 cores/socket)
- 6TB DDR4 Memory (32 DIMM slots) for memory-intensive workloads
- PCIe 4.0 Support for high-speed expansion and low-latency data transfer
2. GPU Acceleration for AI & HPC
- 2U Model: Supports up to 4x GPUs (NVIDIA A100, H100, or AMD Instinct)
- 4U Model: Expands to 8x GPUs for extreme AI training and inferencing
- Optimized airflow and cooling for sustained GPU performance
3. Flexible Storage Options
- 24x 2.5″ or 12x 3.5″ drive bays (mix of NVMe, SAS, SATA)
- High-speed NVMe support (up to 16x in 4U model) for low-latency workloads
- Software-defined storage (SDS) ready for hyperconverged infrastructure
4. Advanced Networking & Expansion
- 100GbE (2U) or 200GbE (4U) networking (OCP 3.0 support)
- Up to 10x PCIe 4.0 slots (4U model) for GPUs, NICs, and accelerators
- NVMe over Fabrics (NVMe-oF) support for high-performance storage networks
5. Enterprise-Grade Reliability
- 80+ Platinum/Titanium PSUs (2000W–2400W) with redundant options
- Lenovo XClarity Controller for simplified management
- ThinkShield Security with TPM 2.0, Secure Boot, and firmware protection
2U vs. 4U Comparison
| Feature | SR670 V2 (2U) | SR670 V2 (4U) |
|---|---|---|
| GPU Support | Up to 4x GPUs | Up to 8x GPUs |
| PCIe Slots | 8x PCIe 4.0 | 10x PCIe 4.0 |
| Storage Bays | 24x 2.5″ or 12x 3.5″ | 24x 2.5″ or 12x 3.5″ |
| Networking | 100GbE standard | 200GbE optional |
| Best For | Virtualization, databases, mid-range AI | Large-scale AI, HPC, GPU-heavy workloads |
Why Choose the SR670 V2?
✅ AI & HPC Ready – Supports NVIDIA HGX & AMD Instinct GPUs
✅ High Storage Density – Flexible NVMe/SAS/SATA configurations
✅ Future-Proof – PCIe 4.0, DDR4, and OCP 3.0 networking
✅ Enterprise Reliability – Redundant power, advanced cooling, and security
Ideal Workloads
- AI/ML Training & Inference
- High-Performance Computing (HPC)
- Virtualization & Private Cloud
- Database & Analytics
- Video Rendering & Simulation
Final Verdict
The ThinkSystem SR670 V2 is a versatile, high-performance server that scales from general enterprise workloads to extreme AI acceleration. The 2U model is perfect for space-constrained environments, while the 4U model excels in GPU-dense deployments.
Best For:
- Businesses needing GPU acceleration
- Data centers requiring high-density compute
- Enterprises future-proofing their infrastructure
Reviews
There are no reviews yet.