Brand: Intel | Category: GPUs
SKU: INTE-HL388 | Part #: HL-388 | MPN: HL-388
Contact for Pricing — Request a Quote
The Intel Gaudi 3 PCIe AI Accelerator (HL-388) from Intel is enterprise-grade GPUs hardware built for data centers. Intel Gaudi 3 AI accelerator in a standard PCIe add-in-card form factor, bringing 1,835 TFLOPS of BF16 compute to any enterprise server. Available in brand new condition. Intel gpus products are trusted by enterprises and data centers worldwide. Available from Omnixon Global. Contact our sales team to request a quote.
| Brand | Intel |
| Category | GPUs |
| SKU | INTE-HL388 |
| Part Number | HL-388 |
| Condition | New |
| Product Line | Intel Gaudi 3 |
| Model / SKU | HL-388 |
| Form Factor | PCIe Add-In Card (FHFL, Double-Slot) |
| Host Interface | PCIe Gen 5 x16 |
| Process Node | TSMC 5nm |
| Tensor Processor Cores (TPCs) | 64 |
| BF16 Compute Performance | 1,835 TFLOPS |
| FP8 Compute Performance | 3,670 TFLOPS |
| INT8 Compute Performance | 3,670 TOPS |
| Memory Type | HBM2e |
| Memory Capacity | 96 GB |
| Memory Bandwidth | 3.7 TB/s |
| Scale-Out Networking | 24 × 200 Gb/s RDMA ports (integrated), via QSFP-DD connectors |
| Scale-Out Fabric Bandwidth | Up to 4.8 Tb/s aggregate bidirectional |
| Thermal Design Power (TDP) | 600 W |
| Cooling | Active (forced-air, onboard fans + server airflow) |
| Power Connectors | PCIe slot + supplemental PCIe power connectors |
| Supported Frameworks | PyTorch, TensorFlow (via SynapseAI SDK and Optimum Habana) |
| Operating System Support | Ubuntu 22.04 LTS, Red Hat Enterprise Linux 8/9 |
| Virtualization Support | SR-IOV capable |
| Management Interface | PCIe management, Intel Gaudi driver stack, Prometheus-compatible telemetry |
| Compliance | RoHS, CE, FCC |