Brand: Huawei | Category: GPUs
SKU: HUAW-ATLAS300IPRO | Part #: Atlas 300I Pro | MPN: Atlas 300I Pro
Contact for Pricing — Request a Quote
The Huawei Atlas 300I Pro Inference Card from Huawei is enterprise-grade GPUs hardware built for data centers. PCIe-form-factor Ascend-powered inference accelerator that slots into standard servers to deliver high-throughput AI inferencing for NLP, vision, and recommendation workloads. Available in brand new condition. Huawei gpus products are trusted by enterprises and data centers worldwide. Available from Omnixon Global. Contact our sales team to request a quote.
| Brand | Huawei |
| Category | GPUs |
| SKU | HUAW-ATLAS300IPRO |
| Part Number | Atlas 300I Pro |
| Condition | New |
| Product Series | Atlas 300I Pro |
| AI Processor | Huawei Ascend (Da Vinci architecture) |
| Form Factor | PCIe add-in card (half-height half-length / full-height configurations) |
| Host Interface | PCIe Gen 4 x16 |
| AI Compute Performance (INT8) | Up to 400 TOPS (INT8) |
| AI Compute Performance (FP16) | Up to 200 TFLOPS (FP16) |
| On-Card Memory Type | HBM2e |
| On-Card Memory Capacity | 32 GB |
| Memory Bandwidth | Up to 1.6 TB/s |
| Supported Precisions | FP32, FP16, BF16, INT8, INT4 |
| Cooling Solution | Active forced-air cooling with onboard fan module |
| Typical Board Power (TDP) | 150 W |
| Power Connector | PCIe auxiliary power (8-pin) |
| Supported Frameworks | MindSpore, TensorFlow, PyTorch, ONNX (via CANN runtime) |
| Software Stack | Huawei CANN (Compute Architecture for Neural Networks), MindX DL inference toolkit |
| Operating System Support | Ubuntu 18.04/20.04, CentOS 7.6/8.2, OpenEuler, Kylin |
| Management Interface | iBMC integration, SNMP, Redfish-compatible host management |
| Operating Temperature | 0 °C to 55 °C (inlet air) |
| Compliance & Certifications | CE, FCC, RoHS, UL |
| Launch Generation | Q2 2025 (Atlas 300I Pro) |