PNY NVIDIA A2
16GB GDDR6 - PCI Express 4.0 x8 (NVIDIA A2)
The NVIDIA A2 GPU features a compact PCIe Gen4 configuration and a configurable thermal envelope (TDP) of 40 to 60 watts, enabling versatile acceleration of inference workflows on any server for deployment at scale.
16GB GDDR6 - PCI Express 4.0 x8 (NVIDIA A2)
NVIDIA A2: AI-powered Graphics and Video Streaming
The NVIDIA A2 GPU enables inference workflows with a compact, high-performance, low-power design for AI applications at the edge. When combined with NVIDIA virtual GPU (vGPU) software, this compact single slot graphics card that consumes only 60W of power can accelerate multiple data centre workloads ranging from graphics rendering on VDI systems to AI applications.from graphics rendering on VDI systems to the most advanced AI applications, all within a secure, versatile and easy-to-manage infrastructure that can adapt to any need.
Powered by the Ampere architecture
- Memory: 16GB GDDR6 (Bandwidth: 200GB/s)
- RT cores: 72
- FP32: 4.5 TFlops
- TF32 Tensor Core: 9 TFlops
- BFLOAT16 Tensor Core: 18 TFlops
- FP16 Tensor Core: 18 TFlops
- INT8 Tensor Core: 36 TOPs
- INT4 Tensor Core: 72 TOPs
- Passive cooling
General information |
Product name |
PNY NVIDIA A2 |
Brand |
PNY | |
Model |
TCSA2MATX-PB | |
Graphics chipset |
Graphics chipset |
NVIDIA A2 |
Number of GPU |
1 | |
Bus |
PCI Express 4.0 x16 (x8) | |
Memory |
Video memory size |
16 GB |
Memory interface |
128 bit(s) | |
Memory type |
GDDR6 | |
Architecture |
Stream processors |
2560 |
Applications |
Multi-GPU |
None |
Use |
Computing station | |
Connectivity |
Video inputs |
None |
Video outputs |
None | |
Design |
Cooling type |
Passive (fanless) |
Consumption |
60 W | |
Low profile |
Yes | |
Format |
Single Slot | |
Power connector(s) |
None | |
Length |
270 mm | |
Warranties |
Commercial warranty |
Seller 3 years |
Legal warranty |
See terms & conditions |
See all PNY products
Product referenced on 26 Jan 2022
For the moment, there are no customer reviews.Write a review