PNY NVIDIA H200 NVL (TCSH200NVLPCIE-PB), 141GB PCIe Hopper Graphics Card
The PNY NVIDIA H200 NVL 141GB PCIe Hopper Graphics Card is a high-performance data centre GPU designed to accelerate demanding AI, deep learning, high-performance computing (HPC) and large-scale data processing workloads. Powered by NVIDIA Hopper architecture, it features 141GB of high-bandwidth GPU memory and delivers up to 4.8TB/s memory bandwidth, making it suitable for handling large language models, generative AI applications and complex computational tasks. Its advanced Tensor Core technology supports multiple precision formats, including FP8, FP16, BFLOAT16 and TF32, enabling efficient AI training and inference. With PCIe Gen5 connectivity, optional NVIDIA NVLink interconnection and Multi-Instance GPU (MIG) support, it offers flexible deployment and scalability for enterprise AI infrastructure. Confidential Computing provides additional data protection, while configurable power consumption of up to 600W supports demanding data centre environments.
Key features
-
NVIDIA Hopper architecture for advanced AI and HPC acceleration.
-
141GB GPU memory with 4.8TB/s memory bandwidth.
-
Up to 3,958 TFLOPS FP8 Tensor Core performance.
-
Supports FP64, FP32, TF32, BFLOAT16, FP16, FP8 and INT8 computing.
-
PCIe Gen5 interface with up to 128GB/s bandwidth.
-
Optional 2-way or 4-way NVIDIA NVLink bridge with 900GB/s interconnect bandwidth.
-
Supports up to 7 Multi-Instance GPUs (MIGs) with 18GB each.
-
Confidential Computing support for enhanced data protection.
Specifications:
| FP64 |
34 TFLOPS |
| FP64 Tensor Core |
67 TFLOPS |
| FP32 |
67 TFLOPS |
| TF32 Tensor Core² |
989 TFLOPS |
| BFLOAT16 Tensor Core² |
1,979 TFLOPS |
| FP16 Tensor Core² |
1,979 TFLOPS |
| FP8 Tensor Core² |
3,958 TFLOPS |
| INT8 Tensor Core² |
3,958 TFLOPS |
| GPU Memory |
141GB |
| GPU Memory Bandwidth |
4.8TB/s |
| Decoders |
7 NVDEC
7 JPEG |
| Confidential Computing |
Supported |
| Max Thermal Design Power (TDP) |
Up to 600W (configurable) |
| Multi-Instance GPUs |
Up to 7 MIGs @ 18GB each |
| Form Factor |
PCIe |
| Interconnect |
2- or 4-way NVIDIA NVLink bridge: 900GB/s
PCIe Gen5: 128GB/s |
| Server Options |
NVIDIA MGX H200 NVL partner and NVIDIA-Certified Systems with up to 8 GPUs |
| NVIDIA AI Enterprise |
Included |
