The NVIDIA A2 Tensor Core GPU features low power consumption, small form factor and high

performance, delivering entry-level inference capabilities for Intelligent Video Analytics (IVA) with

NVIDIA AI deployed at the edge. Designed as a half-height PCIe 4.0 card, this GPU offers configurable

low thermal design power (TDP) of 40–60 watts, providing universal inference acceleration for diverse

servers in large-scale deployments.

Up to 20x Higher Inference PerformanceThe goal of deploying AI inference technology is to create a more convenient life for consumers

through intelligent, real-time experiences. Compared with CPU servers, edge and entry-level servers

equipped with NVIDIA A2 Tensor Core GPUs can deliver up to 20x higher inference performance,

instantly upgrading servers to handle modern AI.

Higher IVA Performance at the Intelligent EdgeIn intelligent edge use cases such as smart cities, manufacturing, and retail, servers powered by the

NVIDIA A2 GPU can deliver up to 1.3x higher performance. Compared to previous-generation GPUs,

the NVIDIA A2 GPU running IVA workloads can improve price-performance and energy efficiency by

up to 1.6x and 10% respectively, thereby enhancing deployment efficiency.

Optimized for Servers The NVIDIA A2 is optimized for inference workloads and deployments in entry-level servers with

constrained space and cooling requirements, such as those in 5G edge and industrial environments.

Offering a half-height form factor that operates within a low-power range, with TDP from 60 watts

down to 40 watts, the A2 is an ideal choice for a wide range of servers.

Industry-Leading AI Inference Performance Across Cloud, Data Center, and Edge AI inference continues to drive breakthrough innovations across industries, including consumer

internet, healthcare and life sciences, financial services, retail, manufacturing, and supercomputing.

The small form factor and low power consumption of the A2, combined with NVIDIA A100 and A30

Tensor Core GPUs, deliver a complete AI inference portfolio across cloud, data center, and edge.

The A2 and NVIDIA AI inference portfolio ensure AI applications are deployed with fewer servers and

less power, resulting in faster insights at significantly reduced cost.

Enterprise-Ready NVIDIA AI Enterprise is an end-to-end cloud-native AI and data analytics software suite, certified to

run on the A2 in virtualized infrastructures based on server virtualization platforms with VMware

vSphere. This enables the management and scaling of AI and inference workloads in hybrid cloud

environments.

Mainstream NVIDIA-Certified Systems NVIDIA-Certified Systems™ with the NVIDIA A2 integrate compute acceleration with high-speed,

secure NVIDIA networking into enterprise data center servers built and sold by NVIDIA’s OEM partners.

Leveraging this program, customers can identify, acquire, and deploy systems on a single

high-performance, cost-effective, and scalable infrastructure to run traditional and diverse modern

AI applications from the NVIDIA NGC™ (NVIDIA GPU Cloud) catalog.