Overview
- NVIDIA's RTX PRO 6000 GPU, part of the Blackwell series, features 96GB of GDDR7 memory, tripling the capacity of the RTX 5090.
- The GPU delivers 125 TFLOPS of FP32 performance, 4000 AI TOPS, and supports a 600W TDP, making it a powerhouse for professional workloads.
- It introduces the new RTX Pro branding, replacing the Quadro and previous RTX naming conventions for professional GPUs.
- The RTX PRO 6000 will be available in workstation, Max-Q, and server variants, with distribution starting in April 2025 and manufacturer availability in May.
- Early benchmarks show a modest 5% improvement in 4K path tracing over the RTX 5090, highlighting its focus on AI and compute-intensive applications.