Overview
NVIDIA A2 Tensor Core GPU delivers 16GB GDDR6 ECC at just 60W in a passive single-slot PCIe 4.0 x16 low-profile card. Designed for edge servers, vGPU virtualization environments, and inference at the lowest power tier of NVIDIA's professional lineup. At 7.6 TFLOPS FP32, throughput is limited — use where power budget and passive integration matter more than raw compute.
Key Features
- 16GB GDDR6 ECC — full ECC protection at 60W TDP
- Passive cooling, low-profile — fits 1U and HTPC servers
- NVIDIA vGPU support for virtualized AI workloads
- PCIe 4.0 x16 (x8 electrical)
- Single-slot form factor
Ideal For
Enterprises deploying AI inference in power-constrained or 1U server environments, and VMware/vSphere administrators who need vGPU support for virtualizing AI workloads across multiple tenants.
Peak FP32
4.5 TF
TF32 Tensor Core
9 TF | 18 TF¹
BFLOAT16 Tensor Core
18 TF | 36 TF¹
Peak FP16 Tensor Core
18 TF | 36 TF¹
Peak INT8 Tensor Core
36 TOPS | 72 TOPS¹
Peak INT4 Tensor Core
72 TOPS | 144 TOPS¹
RT Cores
10
GPU memory
16GB GDDR6
GPU memory bandwidth
200GB/s
$1,365.00
Prices may vary. Verify on vendor site.
Quick Specs
- Peak FP32
- 4.5 TF
- TF32 Tensor Core
- 9 TF | 18 TF¹
- BFLOAT16 Tensor Core
- 18 TF | 36 TF¹
- Peak FP16 Tensor Core
- 18 TF | 36 TF¹
- Peak INT8 Tensor Core
- 36 TOPS | 72 TOPS¹
- Peak INT4 Tensor Core
- 72 TOPS | 144 TOPS¹
Tags
llm-inferencecomputer-visionedge-inferencesmall-model