NVIDIA RTX Pro 6000 WS

Consumer GPU · Blackwell

Summary

The RTX PRO 6000 is a versatile and powerful GPU, with applications from model training to LLM inference to graphics rendering. The workstation edition is slightly more powerful than the server edition in it default configuration, as it has active cooling and higher memory bandwidth, but the RTX PRO 6000 Server Edition is much more widely available among cloud providers and can be configured to have very similar performance in practice.

Launched
Q2 2025
VRAM
96 GB
Mem. bandwidth
1,792 GB/s
On-demand from
$1.29 /GPU/h
Compare prices from 1 providers →

Tech specs

VRAM 96 GB GDDR7
Memory bandwidth 1,792 GB/s
Interface PCIe
CUDA cores 24,064
Tensor cores 752 (Gen 5)
TDP 600 W
Supported data types
FP64FP32TF32FP16BF16FP8FP6FP4INT8

Cloud rental prices

1 provider · available in 3 countries · lowest price per GPU, per hour

Prices updated

On-demand
$1.29 – $1.34 /GPU/h
Spot
$0.53 – $1.34 /GPU/h
Provider On-demand Reserved Spot
Vast.ai $1.29 — $0.53
Compare all NVIDIA RTX Pro 6000 WS cloud providers & configurations →

Models that fit in VRAM

Total VRAM From 16-bit inference 8-bit inference 4-bit inference
1× RTX Pro 6000 WS 96 GB $1.34/h GLM-4.7-FlashQwen3-Coder-30B-A3B-InstructNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 Qwen3-Coder-NextGLM-4.7-FlashNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 DeepSeek-V4-Flashgpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16
2× RTX Pro 6000 WS 192 GB $2.67/h Qwen3-Coder-NextGLM-4.7-FlashNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 DeepSeek-V4-Flashgpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16 MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b
4× RTX Pro 6000 WS 384 GB $5.18/h DeepSeek-V4-Flashgpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16 MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b

Open-weights models that fit in VRAM. Estimated using 🤗 accelerate, plus approximation for up to 8K context.

Media

NVIDIA Blackwell RTX PRO Comes to Workstations and Servers Foster + Partners, Rivian, SoftServe · NVIDIA Newsroom · 18 Mar 2025

Peak theoretical performance

FP4 Tensor Core 2,000 TFLOPS
FP32 125 TFLOPS
RT Core 380 TFLOPS

Performance figures assume no sparsity; in cases where only sparse performance figures are published by the manufacturer, these are halved to give approximate dense performance.

Resources

Detailed documentation from the manufacturer.

Datasheet ↗ Official spec sheet Product page ↗ Manufacturer overview

Similar GPUs

Other accelerators you might compare.

NVIDIA RTX Pro 6000 SE View specs & prices →