Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option
Brand: Lenovo | Category: GPUs
SKU: 4X67A84368 | Part #: 4X67A84368 | MPN: 4X67A84368
Contact for Pricing — Request a Quote
About the Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option
The Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option delivers exceptional compute density at FP32 precision, BF16 for accelerated training, and FP8/INT8 for optimized inference—three critical precision points that define modern AI workloads. Built on the NVIDIA Hopper architecture, this PCIe Gen5 x16 add-in card brings 141 GB of HBM3e memory with 4.8 TB/s memory bandwidth to enterprise data centers, enabling organizations to run generative AI, large language model training and inference, and high-performance computing at scale.
The H200 NVL variant features 4th Generation Tensor Cores that support a full spectrum of precision formats including FP64, FP32, TF32, BF16, FP16, FP8, and INT8—delivering flexibility across scientific computing and AI applications. With integrated active cooling and a 700 W thermal design power envelope, the Lenovo model 4X67A84368 integrates seamlessly into Lenovo ThinkSystem servers designed for demanding workloads. NVLink connectivity enables multi-GPU scaling for enterprises requiring distributed compute. For IT infrastructure teams and AI platform engineers evaluating next-generation accelerators, this GPU option represents a validated, vendor-supported solution across UAE, GCC, EMEA, and APAC regions.
Key Specifications
- Manufacturer: Lenovo
- Part Number: 4X67A84368
- GPU Model: NVIDIA H200 NVL
- Memory Capacity: 141 GB HBM3e
- Memory Bandwidth: 4.8 TB/s
- PCIe Interface: PCIe Gen5 x16
- Thermal Design Power: 700 W
- Supported Precisions: FP64, FP32, TF32, BF16, FP16, FP8, INT8
Contact Omnixon Global to request a quote for the Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option and explore integration options for your enterprise infrastructure.
Technical Specifications
| Brand | Lenovo |
| Category | GPUs |
| SKU | 4X67A84368 |
| Part Number | 4X67A84368 |
| Condition | New |
| Manufacturer Part Number | 4X67A84368 |
| Product Name | ThinkSystem NVIDIA H200 NVL PCIe GPU Option |
| GPU Architecture | NVIDIA Hopper |
| GPU Model | NVIDIA H200 NVL |
| Form Factor | PCIe Add-in Card |
| Memory Type | HBM3e |
| Memory Capacity | 141 GB |
| Memory Bandwidth | 4.8 TB/s |
| PCIe Interface | PCIe Gen5 x16 |
| Tensor Core Generation | 4th Generation (Hopper) |
| Supported Precisions | FP64, FP32, TF32, BF16, FP16, FP8, INT8 |
| NVLink Support | NVLink (NVL variant; PCIe host interface) |
| Cooling | Active (integrated fan cooling) |
| Thermal Design Power (TDP) | 700 W |
| Product Series | Lenovo ThinkSystem GPU Options |
| Target Platform | Lenovo ThinkSystem Servers |
| Intended Workloads | Generative AI, LLM Training and Inference, HPC, Scientific Computing |
| Regional Availability | UAE, GCC, EMEA, APAC |
Frequently Asked Questions about Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option
What does the Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option do?
The Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option accelerates AI/ML training, inference, scientific HPC, and virtualization (vGPU) workloads. Typical deployments include LLM training clusters, computer-vision pipelines, financial risk modeling, and rendering farms.
What are the headline specs of the Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option?
Key specifications for the Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option: new condition; raid support H200; gpu support NVIDIA H200; ai optimized Yes. Manufacturer part number 4X67A84368. For the full datasheet with electrical, environmental, and compliance details, contact our pre-sales engineering team.
Is the Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option compatible with my infrastructure?
The Lenovo ThinkSystem NVIDIA H200 NVL PCIe GPU Option requires a PCIe Gen4 or Gen5 x16 slot, server power adequate for the card's TDP, and CUDA/ROCm driver support in your hypervisor or bare-metal OS. Sales engineering will confirm chassis fit (1U/2U/4U), PCIe lane count, and PSU headroom before quoting.