Lenovo ThinkSystem SR670 V3 with Intel Gaudi 3 OAM 128GB GPU
Brand: Lenovo | Category: GPUs
SKU: 7D9DA015NA | Part #: 7D9DA015NA | MPN: 7D9DA015NA
Contact for Pricing — Request a Quote
About the Lenovo ThinkSystem SR670 V3 with Intel Gaudi 3 OAM 128GB GPU
AI training and large-scale inference workloads demand purpose-built acceleration architecture, and the Lenovo ThinkSystem SR670 V3 with Intel Gaudi 3 OAM 128GB GPU delivers exactly that in a compact 2U form factor. This server consolidates eight Intel Gaudi 3 OAM accelerator modules—each equipped with 128GB of HBM2e memory—into a single system capable of 1TB total accelerator memory when fully populated. Built on PCIe 5.0 infrastructure with high-speed Intel Gaudi 3 scale-up fabric for intra-node interconnect, the SR670 V3 is engineered for AI infrastructure teams and enterprise data center architects seeking dense, power-efficient GPU compute without compromise.
The Lenovo ThinkSystem SR670 V3 pairs dual-socket support for Intel Xeon Scalable Processors (4th or 5th Generation) with 32 DDR5 memory DIMM slots, ensuring CPU and memory bandwidth scale alongside accelerator throughput. Flexible storage connectivity via NVMe and SATA/SAS controllers, combined with high-speed Ethernet and InfiniBand expansion through PCIe, positions this platform for multi-node cluster deployments. Redundant hot-swap power supplies and choice of air cooling or Direct Water Cooling (DWC) provide operational resilience and thermal flexibility for continuous AI workloads. Lenovo XClarity Controller management and compatibility with Lenovo XClarity Administrator streamline provisioning and monitoring across your fleet. Red Hat Enterprise Linux, Ubuntu, and SUSE Linux Enterprise Server support round out a production-ready stack.
Key Specifications
- Manufacturer: Lenovo
- Part Number: 7D9DA015NA
- Form Factor: 2U Rack Server
- AI Accelerator: Intel Gaudi 3 OAM; up to 8 modules; 128GB HBM2e per module; 1TB total memory when fully populated
- Processor Support: Intel Xeon Scalable Processors (4th or 5th Generation); 2 sockets
- Memory: 32 DDR5 DIMM slots
- Connectivity: PCIe 5.0; high-speed Ethernet and InfiniBand via expansion
- Cooling: Air cooling and Direct Water Cooling (DWC) options
Contact Omnixon Global to request a quotation for the Lenovo ThinkSystem SR670 V3 (7D9DA015NA) and discuss deployment requirements for your AI infrastructure.
Technical Specifications
| Brand | Lenovo |
| Category | GPUs |
| SKU | 7D9DA015NA |
| Part Number | 7D9DA015NA |
| Condition | New |
| Manufacturer Part Number | 7D9DA015NA |
| Product Line | ThinkSystem SR670 V3 |
| Form Factor | 2U Rack Server |
| AI Accelerator | Intel Gaudi 3 OAM |
| Accelerator Memory per Module | 128GB HBM2e |
| Maximum Accelerator Modules | 8x Intel Gaudi 3 OAM |
| Total Accelerator Memory (fully populated) | 1TB HBM2e (8 × 128GB) |
| Processor Support | Intel Xeon Scalable Processors (4th or 5th Generation) |
| Processor Sockets | 2 |
| Memory Type | DDR5 |
| Memory Slots | 32 DIMM slots |
| Storage Interface | NVMe and SATA/SAS support via onboard controllers |
| Network Connectivity | Supports high-speed Ethernet and InfiniBand via PCIe expansion |
| PCIe Generation | PCIe 5.0 |
| Power Supply | Redundant hot-swap power supplies |
| Cooling Options | Air cooling and Direct Water Cooling (DWC) |
| Management | Lenovo XClarity Controller (XCC), compatible with Lenovo XClarity Administrator |
| Operating System Support | Red Hat Enterprise Linux, Ubuntu, SUSE Linux Enterprise Server |
| Accelerator Interconnect | Intel Gaudi 3 scale-up high-speed fabric (intra-node) |
| Chassis | Standard 19-inch rack compatible, 2U |
Frequently Asked Questions about Lenovo ThinkSystem SR670 V3 with Intel Gaudi 3 OAM 128GB GPU
What server platforms accept the Lenovo ThinkSystem SR670 V3 with Intel Gaudi 3 OAM 128GB GPU?
Reference servers include Dell PowerEdge XE9680 / XE9712, HPE Cray XD670, Lenovo ThinkSystem SR685a / SR675 V3, Supermicro AS-A21GE / SYS-821GE, Gigabyte G593 / G894, ASUS ESC. Share your target platform in the RFQ and we will confirm chassis-to-GPU compatibility and recommended NIC pairing.
How long is the lead time on AI GPUs?
Highly model-dependent. L40S / RTX-class: typically 3-6 weeks. H100/H200/B200 in SXM form factor: 12-16 weeks for whole-platform allocations. We quote genuine-channel ETAs only — no grey-market promises.
Do you supply matched networking (Quantum InfiniBand / Spectrum-X)?
Yes — Omnixon stocks the full NVIDIA networking lineup (Quantum-2 / Quantum-X InfiniBand, Spectrum-X Ethernet, ConnectX NICs, BlueField DPUs) so we can quote a complete training-cluster BOM, not just the GPUs.
Can you help with NVIDIA AI Enterprise licensing?
Yes. We hold genuine-channels for NVIDIA AI Enterprise software subscriptions. Add it to your RFQ and we quote node-aligned licensing along with the hardware.