Brand: NVIDIA | Category: GPUs
SKU: NVID-H100PCIE80GB | Part #: H100-PCIE-80GB | MPN: H100-PCIE-80GB
Contact for Pricing — Request a Quote
The NVIDIA H100 PCIe 80GB HBM2e GPU Accelerator from NVIDIA is enterprise-grade GPUs hardware built for data centers. Enterprise-class 80GB HBM2e GPU accelerator in standard PCIe form factor for large-scale AI inference and HPC workloads across on-premises data centers. Available in brand new condition. NVIDIA gpus products are trusted by enterprises and data centers worldwide. Available from Omnixon Global. Contact our sales team to request a quote.
| Brand | NVIDIA |
| Category | GPUs |
| SKU | NVID-H100PCIE80GB |
| Part Number | H100-PCIE-80GB |
| Condition | New |
| Product Line | H100 PCIe |
| Architecture | Hopper |
| GPU Memory | 80GB HBM2e |
| Memory Interface | 5120-bit |
| Memory Bandwidth | 3.35 TB/s |
| CUDA Cores | 14080 |
| Tensor Cores | 880 |
| Max Power Consumption | 350W |
| Form Factor | PCIe Gen5 Standard Profile |
| Host Interface | PCI Express 5.0 (x16) |
| Peak FP32 Performance | 67.1 TFlops |
| Peak FP8 Performance | 1073.8 TFlops |
| Peak Tensor (sparsity-enabled) | 1457.2 TFlops |
| Manufacturing Process | 5nm TSMC |
| Cooling | Passive (requires server airflow) |
| Software Support | NVIDIA CUDA 12.x, cuDNN, TensorRT, Triton Inference Server |
| Operating Temperatures | 0°C to 60°C |
| Physical Dimensions | 267 x 111 x 42 mm |
| Typical Server Integration | 1-8 units per 2U/3U enterprise server chassis |