NVIDIA RTX PRO 6000 Blackwell Max-Q Edition 96GB GDDR7

The NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition is a high-efficiency professional GPU based on the Blackwell platform, designed for compact, high-density multi-GPU workstation deployments.It features 96GB high-speed GDDR7 memory with ECC error correction, fifth-generation Tensor Cores and fourth-generation Ray Tracing Cores. Rated at 300W TDP, it delivers outstanding power efficiency compared to the standard 600W RTX PRO 6000 workstation variant, enabling up to four GPUs installed within one workstation chassis.

It excels at local LLM inference, agentic AI development, small-batch model fine-tuning, data science, photorealistic rendering, real-time simulation, Omniverse digital twins and high-end visualisation workloads. The large 96GB ECC memory allows huge AI models to run fully on local hardware without cloud offloading.This card supports MIG (Multi-Instance GPU), splitting the GPU into up to four isolated 24GB instances for multi-user or multi-task workload isolation. It adopts PCIe 5.0 x16 interface and is equipped with four DisplayPort 2.1b outputs for multi 8K/16K professional monitor setups.

It is the preferred choice for AI research labs, system integrators and engineering teams building dense multi-GPU local AI workstations where power consumption and thermal density are critical constraints.