Ampere GPUs

Data Center GPU for AI training, inference, data analytics and high-performance computing

NVIDIA A2 is a low-power, entry-level GPU designed for inference at the edge, suitable for AI and video analytics applications in data centers.

NVIDIA A10 provides versatile GPU acceleration for AI, rendering, and VDI, enabling high-performance computing across enterprise data centers.

NVIDIA-A30 is an NVIDIA A30 Tensor Core GPU for data center AI inference, enterprise training, and HPC. It provides 24GB HBM2 memory, 933GB/s bandwidth, PCIe Gen4, and 165W TDP. View the official Datasheet and request current pricing or a quote.

NVIDIA A40 is a versatile GPU designed for AI, HPC, and graphics workloads, offering high-end performance for data centers.

NVIDIA A16 is a GPU optimized for virtual desktop infrastructure, providing high-efficiency performance for cloud and AI applications.

NVIDIA A100-40G delivers efficient performance for AI inference and training, optimized for data centers with mid-range computational needs.

NVIDIA A100-80G is a powerful GPU designed for AI, data analytics, and scientific computing in modern data centers.

A800-80G is an NVIDIA A800 80GB data center GPU for AI training, data analytics, and high-performance computing. It features 80GB HBM2 memory, 8,192 FP32 CUDA Cores, 512 third-generation Tensor Cores, MIG partitioning, and NVLink-based multi-GPU scaling.