MPN: GPU-NVH200NVL·By: Supermicro
Ships from the US.
We'll confirm availability and pricing for GPU-NVH200NVL, usually within one business day.
We'll confirm availability and pricing for GPU-NVH200NVL, usually within one business day.
The Supermicro GPU-NVH200NVL is a PCIe 5.0 x16 graphics card built around the NVIDIA H200 NVL Tensor Core GPU, designed for data center acceleration and high-performance workloads. This model integrates 141GB of HBM3e GPU memory, enabling large in-memory datasets and complex models to reside closer to the processor for reduced latency and improved throughput. The card uses an air-cooled design, supporting deployments in compatible server platforms that are optimized for airflow-based thermal management. As a PCIe 5.0 x16 device, it can be integrated into modern server architectures that require high-bandwidth connectivity to host CPUs and other peripherals. The large HBM3e memory capacity and Tensor Core GPU architecture make this card suitable for environments that need substantial compute density and memory resources within a standard PCIe form factor, supporting scalable, rack-based data center configurations.
| Manufacturer part # | GPU-NVH200NVL |
|---|---|
| Brand | Supermicro |
| Condition | New |
| Model | GPU-NVH200NVL |
| GPU Model | NVIDIA H200 NVL Tensor Core GPU |
| Part Class | Supermicro branded NVIDIA data center GPU option |
| Form Factor | PCIe add-in card |
| Memory Type | HBM3e |
| Cooling Type | Air cooled |
| Product Line | NVIDIA H200 NVL |
| Product Type | Graphics Processing Unit |
| Tensor Cores |
Suitable for data center environments that require a PCIe 5.0 x16 graphics card built on the NVIDIA H200 NVL Tensor Core GPU, supporting high-performance compute workloads with 141GB of HBM3e memory.
Ideal for servers that provide PCIe 5.0 x16 slots, enabling high-bandwidth connectivity between the GPU-NVH200NVL accelerator and host systems in modern compute platforms.
A good fit for workloads that benefit from large on-board memory, leveraging the 141GB of HBM3e GPU memory to keep expansive datasets and models resident on the GPU for improved efficiency.
Designed for systems using air cooling, allowing deployment of the NVIDIA H200 NVL Tensor Core GPU in data center racks that rely on airflow-based thermal management rather than liquid cooling.
Appropriate for applications that can benefit from a Tensor Core GPU architecture, using the NVIDIA H200 NVL Tensor Core GPU as an accelerator in compatible server configurations.
Details
Shipping, warranty & returns
No reviews yet — be the first to review this part.
Secure SSL checkout. Visa, Mastercard, Amex and Discover accepted. ACH / wire available for businesses.
New, certified-refreshed and OEM stock with guaranteed compatibility, backed by warranty on every part.
We process quotes, POs, stock documentation and compliance invoicing.
[email protected]| 528 |
| Graphics Chip | NVIDIA H200 |
|---|
| Memory Capacity | 141 GB |
|---|
| Stream Processors | 16896 |
|---|
| PCI Express Version | PCIe Gen 5.0 |
|---|
| Theoretical FP8 Performance | 3958 TFLOPS |
|---|
| Theoretical FP32 Performance | 67 TFLOPS |
|---|
| PCI Express Lane Configuration | x16 |
|---|
| Bus Interface | PCIe 5.0 x16 |
|---|
The Supermicro GPU-NVH200NVL is built around the NVIDIA H200 NVL Tensor Core GPU. This means it is a Tensor Core GPU accelerator, suitable for environments that plan to standardize on NVIDIA H200 NVL based compute resources and want a PCIe 5.0 x16 add-in card form factor for their systems.
This graphics card provides 141GB of HBM3e GPU memory. The large 141GB capacity is beneficial for workloads that require substantial in-memory datasets or large models, allowing more data to remain resident on the GPU and reducing the need for frequent transfers between system memory and the accelerator.
The GPU-NVH200NVL uses a PCIe 5.0 x16 interface. To deploy this accelerator, the host system must provide a compatible PCIe 5.0 x16 slot, ensuring sufficient bandwidth between the GPU and the CPU and aligning with modern server platforms that support PCIe 5.0 expansion cards.
This model is specified as air cooled. It is intended for installation in server or workstation platforms that provide adequate airflow-based cooling, allowing data center and IT teams to integrate the GPU into existing air-cooled racks without requiring liquid cooling infrastructure.
The GPU-NVH200NVL is categorized as a GPU, specifically a PCIe 5.0 x16 graphics card based on the NVIDIA H200 NVL 141GB HBM3e Tensor Core GPU. This classification helps procurement and engineering teams align it with accelerator standards and ensure it fits within their PCIe GPU deployment strategy.
You might also want to check out these products.

900-2H403-0000-000 - Nvidia Tesla P100 16GB HBM2 NVlink SXM2 GPU

900-22081-0040-000 - NVIDIA Tesla K40 Video Graphic Card

100-506115 - AMD Radeon Pro WX 3200 Graphic Card 4GB GDDR5

900-1G611-2530-000 - Nvidia TITAN Xp 12GB GDDR5X Graphics Card

900-52055-0310-000 - NVIDIA Tesla Grid K2 8GB Video Graphics Card

900-2G183-0000-001 - Nvidia Tesla T4 Turing 16GB GDDR6 PCIe x16 GPU

699-2G402-0060-300 - NVIDIA Tesla M60 16GB GDDR5 Video Graphics Card

900-22081-0110-000 - NVIDIA Tesla K20 5GB GDDR5 320-Bit GPU