Intel Gaudi 2 AI Accelerator OAM Module 96GB HBM2e
Brand: Intel | Category: GPUs
SKU: HLS-GAUDI2-OAM | Part #: HLS-GAUDI2-OAM | MPN: HLS-GAUDI2-OAM
Contact for Pricing — Request a Quote
About the Intel Gaudi 2 AI Accelerator OAM Module 96GB HBM2e
The Intel Gaudi 2 AI Accelerator OAM Module delivers exceptional compute density for enterprise AI workloads, with 24 Tensor Processor Cores built on a 7nm process node. This OCP Accelerator Module (OAM) form factor integrates 96 GB of HBM2e memory across 6 stacks, paired with 2.45 TB/s memory bandwidth—critical for training and inference tasks requiring sustained high-throughput data access. The module features 24 × 100GbE RoCE v2 RDMA on-die networking ports, enabling 2.4 Tb/s bidirectional interconnect bandwidth for distributed multi-accelerator configurations. Host connectivity is provided via PCIe Gen 4 x16, allowing seamless integration into existing server architectures. Part number HLS-GAUDI2-OAM supports FP32, BF16, FP16, and INT8 precision formats, making it adaptable to diverse AI inference and training scenarios.
Data center AI infrastructure teams and enterprise procurement organizations evaluating next-generation accelerator platforms will find the Intel Gaudi 2 compelling for deep learning, large language models, and computer vision workloads. The OAM specification enables flexible deployment—up to 8 modules per Universal Baseboard server node—without proprietary vendor lock-in, while the Intel SynapseAI SDK and Intel Gaudi Software stack provide framework compatibility with PyTorch and TensorFlow. Linux environments (Ubuntu, CentOS/RHEL) are fully supported, reducing integration friction for existing deployments. Organizations seeking cost-effective, high-performance alternatives to proprietary GPU accelerators benefit from the module's balance of memory capacity, interconnect bandwidth, and open standards compliance.
Omnixon Global stocks the Intel Gaudi 2 AI Accelerator OAM Module and supports organizations across the Middle East and beyond with pre-sales engineering, integration guidance, and procurement at enterprise scale. To request a quote, technical datasheet, or configuration assistance for the HLS-GAUDI2-OAM, please submit an RFQ through Omnixon Global's B2B portal or contact your dedicated account team today.
Technical Specifications
| Brand | Intel |
| Category | GPUs |
| SKU | HLS-GAUDI2-OAM |
| Part Number | HLS-GAUDI2-OAM |
| Condition | New |
| Manufacturer Part Number | HLS-GAUDI2-OAM |
| Product Family | Intel Gaudi 2 |
| Form Factor | OAM (OCP Accelerator Module) |
| Process Node | 7nm |
| Tensor Processor Cores (TPCs) | 24 |
| Memory Type | HBM2e |
| Total Memory Capacity | 96 GB |
| Memory Bandwidth | 2.45 TB/s |
| HBM2e Stacks | 6 |
| On-Die Networking Ports | 24 × 100GbE RoCE v2 RDMA |
| Total On-Die Networking Bandwidth | 2.4 Tb/s bidirectional |
| Host Interface | PCIe Gen 4 x16 |
| Supported AI Frameworks | PyTorch, TensorFlow (via SynapseAI SDK) |
| Software Stack | Intel SynapseAI SDK, Intel Gaudi Software |
| OAM Compliance | OCP Accelerator Module (OAM) specification |
| Typical Server Density | Up to 8 OAM modules per Universal Baseboard server node |
| Supported Precision Formats | FP32, BF16, FP16, INT8 |
| Target Workloads | AI training and inference, deep learning, LLM, computer vision |
| Compatible Operating Systems | Linux (Ubuntu, CentOS/RHEL) |
Frequently Asked Questions about Intel Gaudi 2 AI Accelerator OAM Module 96GB HBM2e
What server platforms accept the Intel Gaudi 2 AI Accelerator OAM Module 96GB HBM2e?
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.