Designed for compatible HPE ProLiant server environments.
PCIe accelerator with four integrated GPUs in a server expansion form factor.
Provides 64GB of GDDR6 memory for high-density VDI workloads.
The NVIDIA A16 64GB PCIe Non-CEC Accelerator for HPE provides GPU acceleration for high-density virtual desktop infrastructure and remote workstation environments. It uses the NVIDIA Ampere architecture and combines four GPUs on one PCIe accelerator.
The card includes 64GB of GDDR6 memory, with 16GB allocated to each GPU. Each GPU provides memory bandwidth of 200 GB/s, delivering a combined 4 x 200 GB/s memory bandwidth configuration for graphics-intensive and compute-enabled workloads.
Part number R8T26C is designed for compatible HPE ProLiant server platforms. When paired with NVIDIA Virtual PC software, the A16 supports centralized virtual desktops and remote graphics applications from a server-based infrastructure.
| Part Number | R8T26C |
| Manufacturer | NVIDIA |
| Product Type | GPU accelerator |
| Model | A16 |
| Architecture | NVIDIA Ampere |
| Form Factor | PCIe accelerator card |
| Interface | PCIe |
| Memory Capacity | 64GB |
| Memory Type | GDDR6 |
| GPU Configuration | 4 GPUs, 16GB per GPU |
| Memory Bandwidth | 4 x 200 GB/s |
The NVIDIA A16 is a four-GPU PCIe accelerator built for high-density virtual desktop infrastructure, remote workstations, and server-based graphics applications.
The accelerator provides 64GB of GDDR6 memory in total. Each of its four GPUs includes 16GB of memory.
Each GPU provides 200 GB/s of memory bandwidth, represented as a 4 x 200 GB/s configuration across the accelerator.
The NVIDIA A16 64GB PCIe Non-CEC Accelerator is designed for compatible HPE ProLiant server platforms.
The A16 is designed for high-density VDI, virtual desktops, remote workstations, graphics-intensive applications, and selected HPC workloads.
Yes. The A16 works with NVIDIA Virtual PC software to provide centralized virtual desktop and remote graphics capabilities.