A30

The NVIDIA A30 features the Ampere architecture with 24 GB of HBM2e VRAM. It is a compact datacenter accelerator designed for scalable AI inference and mainstream training workloads, offering HBM2e memory bandwidth in a power-efficient form factor.

Its 24 GB of VRAM supports Large Language Models (LLM) and fine-tuning of mid-sized models with high memory bandwidth. The A30 is well-suited for inference-heavy deployments where organizations need to maximize GPU density and throughput per watt across datacenter racks.