Nvidia Launches RTX PRO 5500 Blackwell Workstation GPU With 84GB GDDR7

Nvidia has released the RTX PRO 5500 Blackwell professional workstation graphics card. The card uses the Blackwell architecture and is equipped with 84GB of GDDR7 memory. The memory supports error-correcting code (ECC) and delivers bandwidth of 1,398 GB/s. This configuration can run larger AI models, support longer context windows, and run multiple models simultaneously without offloading to system memory.

On the core architecture side, the RTX PRO 5500 is built on NVIDIA Blackwell and carries fifth-generation Tensor Cores and fourth-generation RT Cores. Blackwell's SM units introduce Neural Shaders, which integrate neural networks inside programmable shaders. The workstation version has three NVENC and three NVDEC decoders, provides up to four DisplayPort 2.1b interfaces, has a maximum power draw of 600W, and is offered with active air-cooling and liquid-cooling solutions.

The RTX PRO 5500 supports Multi-Instance GPU (MIG) technology, which can divide a single GPU into multiple fully isolated instances. One instance can use up to 84GB of memory, or two instances can each use up to 42GB. Each instance has its own high-bandwidth memory, cache and compute cores, and quality of service (QoS) is guaranteed. The product is designed for rack-mounted workstation deployments, used to accelerate agentic AI, physical simulation and graphics workloads.