Be the first to review “NVIDIA® A2” Cancel reply
Our website uses cookies to give you the best experience possible. By continuing to use our site, you agree with our use of cookies outlined in our Privacy Policy
NVIDIA® A2
SKU: NVBGPUNTA2
Categories: NVIDIA Data Centre GPUs
If you need a low-power, high-performance GPU for inference at the edge, the NVIDIA A2 Tensor Core GPU is the perfect choice. It is a compact PCIe Gen4 card that can fit in any server and run on 40–60 W of power. It can handle intelligent video analytics (IVA) or NVIDIA AI at the edge with ease.
The A2 GPU is ideal for scaling up edge inference without breaking the bank. It can boost inference performance by up to 20 times compared to CPUs and improve IVA efficiency by 1.3 times compared to previous GPUs. It also works with NVIDIA-Certified systems and NVIDIA AI software, including the NVIDIA Triton Inference Server, to deliver fast and reliable inference across edge, data center, and cloud. With the A2 GPU, you can deploy AI applications with fewer servers and less power, saving time and money.
Details & Specifications
Details & Specifications
| Manufacturer Part Number | TCSA2ATX-PB |
|---|---|
| Manufacturer |
Additional:
| CODECs Supported | H.265 | H.265| VP9 | AV1 Decode |
|---|---|
| Operating Humidity | 5% to 85% Relative Humidity |
| Operating Temperature | 0 to 50 degrees Centigrade |
| Video Encode | Decode | 1x Video Encoder | 2x Video Decoder |
Cooling:
| Hardware Root of Trust | Yes |
|---|---|
| NEBS ready | Level 3 |
| Passive Cooling | Yes |
Dimensions:
| Form Factor | Low-Profile PCIe | Single Slot |
|---|---|
| Item Weight (Kg) | 6.61” L x 2.71” H |
| NVIDIA CUDA Support | 11.1 or later |
| vGPU Software Revision | 14.0 or later |
GPU:
| # CUDA Cores | 1280 |
|---|---|
| # RT Cores | 10 | Gen 2 |
| ECC Support | Yes |
| Game Clock | 1440 MHz Base | 1770 MHz Boost |
| GPU Architecture | |
| GPU Slot Size | |
| Lithography | 7nm |
| Peak FP16 Tensor Core Performance (TFLOPS) | 36 | Sparsity |
| Peak FP32 Performance (TFLOPS) | 4.5 |
| Peak FP32 Tensor Core Performance (TFLOPS) | 18 | Sparsity |
| Peak Int4 Performance (TOPS) | 72 | Sparsity |
| Peak Int8 Performance (TOPS) | 144 | Sparsity |
Memory:
| Memory Bandwidth (GB/s) | 200 |
|---|---|
| Memory Capacity (GB) | 16 |
| Memory Frequency (MHz) | 6251 |
| Memory Interface (bit) | 128 |
| Memory Type | GDDR6 ECC |
PCIe:
| PCIe Version | PCIe 4.0 |
|---|
Power:
| Max Power Consumption (W) | 40-60 | Configurable |
|---|
Software:
| Supported OS | Linux, Windows |
|---|
Warranty:
| Mfr Warranty Length (M) | 36 |
|---|---|
| Mfr Warranty Type | Direct To Manufacturer |
Reviews
There are no reviews yet.