NVIDIA A2 Tensor Core

NVIDIA A2 Tensor Core

NVIDIA A2 Tensor Core GPU: Entry-level acceleration for demanding workloads

NVIDIA A2 Tensor Core

We are delighted to introduce the NVIDIA A2 Tensor Core GPU - your gateway to powerful AI inferencing and graphics for any server. This compact and efficient powerhouse delivers 20x faster performance than CPUs, unlocking new possibilities for your business or data center.

Key Features

  • Powerful Tensor Cores - The A2 utilises NVIDIA's Tensor Cores, specifically designed for accelerating AI workloads and delivering efficient performance for tasks like deep learning training and inference.
  • Broad framework support - NVIDIA's A2 GPU seamlessly integrates with all major AI frameworks, including NVIDIA Triton Inference Server, TensorFlow, and others. This allows for flexibility and compatibility when deploying AI applications without needing to rewrite code for different frameworks.
  • Scalable inference performance - This powerhouse GPU offers scalable inference performance, enabling businesses to start with a single card and scale up by adding more A2 GPUs to their servers as their needs grow. This provides a cost-effective way to handle increasing workloads without needing to invest in entirely new hardware infrastructure.

Specifications

Architecture Ampere
GPU memory 16GB GDDR6
Memory Bandwidth 200GB/s
Peak FP32 4.5 TF
TF32 Tensor Core 9 TF || 18 TF with sparsity
BFLOAT16 Tensor Core 18 TF || 36 TF with sparsity
Peak FP16 Tensor Core 18 TF || 36 TF with sparsity
Peak INT8 Tensor Core 36 TOPS || 72 TOPS with sparsity
Peak INT4 Tensor Core 72 TOPS || 144 TOPS with sparsity
RT Cores 10
Media Engines 1 * video decoder
2 * video decoders
Includes AV1 decode
Interconnect PCIe Gen4 x8
Form Factor 1-slot, LP PCIe
Max Thermal Design Power (TDP) 40-60W (Configurable)
vGPU Support NVIDIA Virtual PC (vPC)
NVIDIA Virtual Applications (vApps)
NVIDIA RTX Virtual Workstations (vWS)
NVIDIA AI Enterprise
NVIDIA Virtual Compute Server (vCS)