RTX PRO 6000 Blackwell Server

The RTX PRO 6000 Blackwell Server is an NVIDIA GPU built on the Blackwell 2.0 architecture, released 2025-03-18. For running AI locally, the numbers that matter are its 96 GB of GDDR7 and 1790 GB/s of memory bandwidth. VRAM decides which models fit at all; bandwidth sets how fast they generate text.

What you can run on 96 GB

At Q4_K_M quantization (the usual local default), 96 GB holds models up to roughly 152B parameters, leaving headroom for context. On this card you can run, among others:

• Qwen3.5 122B A10B - 125.1B parameters
• gpt-oss-120b - 120.4B parameters
• Qwen3-Next-80B-A3B-Instruct - 81.3B parameters
• Qwen3-Coder-Next - 79.7B parameters
• Qwen2.5 72B - 72B parameters

Larger models need a higher-VRAM card, a second GPU, or CPU offload (which is much slower). Check any specific model with the VRAM calculator, or see the full picture on what can I run.

Local LLM speed (LLaMA 3, llama.cpp)

Single-stream token-generation throughput - estimated from memory bandwidth:

Model (quant)	Speed on RTX PRO 6000 Blackwell Server
Llama 3 8B (Q4_K_M)	209.5 tok/s
Llama 3 8B (F16)	95.1 tok/s
Llama 3 70B (Q4_K_M)	24.6 tok/s

Because decode is memory-bandwidth bound, the 1790 GB/s figure is the best single predictor of chat speed on this card. Estimates are calibrated against measured RTX-40-series cards and are typically within ~15%.

Memory and power

• VRAM: 96 GB GDDR7 (512-bit bus)
• Bandwidth: 1790 GB/s
• TDP: 600 W - a 1000 W+ power supply is recommended
• Process: 5 nm
• Interface: PCIe 5.0 x16

Quantization and context

Quantization trades a little quality for a lot of VRAM. On 96 GB you can fit roughly a 152B model at Q4_K_M, about a 82B model at the higher-quality Q8, or a smaller model at full FP16. Longer context windows also consume VRAM (the KV cache grows with context length), so leave a few GB of headroom if you plan to use large prompts or many concurrent requests. For most chat and coding use, Q4_K_M on this card is the sweet spot between speed, quality, and the 96 GB budget.

How it compares

Similar cards for local AI, by VRAM and 8B-Q4 speed:

GPU	VRAM	Bandwidth	Llama 3 8B Q4
RTX PRO 6000 Blackwell Server	96 GB	1790 GB/s	209.5 tok/s
RTX PRO 6000 Blackwell	96 GB	1790 GB/s	209.5 tok/s
RTX PRO 6000 Blackwell Max-Q	96 GB	1790 GB/s	209.5 tok/s
RTX PRO 5000 72 GB Blackwell	72 GB	1340 GB/s	156.8 tok/s

Bottom line

The RTX PRO 6000 Blackwell Server is best for llm-inference, image-gen, comfyui. With 24 GB+ it comfortably handles most open models, including 30B-class at Q4. If you need more, compare with RTX PRO 6000 Blackwell and RTX PRO 6000 Blackwell Max-Q.

Sources

• Specifications: RightNow GPU Database (TechPowerUp data)
• Benchmarks: GPU-Benchmarks-on-LLM-Inference (basis for the bandwidth estimate)

Specs and benchmarks last checked 2026-06-08. Verify current pricing before buying.