(858) 613-1818Quote list 0

NVIDIA RTX PRO 6000 Blackwell Server Edition

The Blackwell data center PCIe GPU with 96 GB GDDR7 ECC, 24,064 CUDA cores, 4 PFLOPS FP4 and 400 to 600 W, for air cooled servers with up to ten GPUs.

Family
Blackwell
Edition
PCIe server card, passive
Memory
96 GB GDDR7 with ECC per GPU
Memory bandwidth
1.6 TB/s per GPU
Interconnect
PCIe Gen5 x16, no NVLink
Form factor
Dual slot full height full length air cooled card, or single slot liquid cooled variant
TDP
400 W to 600 W (configurable)
Cooling
Passive
Best for
Dense PCIe inference, fine tuning, rendering and virtual workstations from 1 to 10 GPUs per server
Availability
Build to order

RTX PRO 6000 Blackwell Server Edition is the passively cooled, data center version of NVIDIA's Blackwell workstation GPU. It carries 96 GB of GDDR7 with ECC at 1.6 TB/s, 24,064 CUDA cores, fifth generation Tensor cores with 4 PFLOPS of FP4, and 188 fourth generation RT cores. Power is configurable between 400 W and 600 W, and the card supports Multi Instance GPU with up to four isolated instances.

Because it is a dual slot PCIe card without NVLink, it suits servers that pack two to ten GPUs for inference, fine tuning of models up to roughly 70B parameters with tensor parallelism over PCIe, media and rendering farms, and virtual workstations. Four NVENC and four NVDEC engines make it strong for video pipelines. We supply it in 2U, 4U and 5U PCIe servers and as an individual card, subject to end user registration.

Systems that carry this GPU

Related guides

Buying data center GPUs, export and registration

The compliance steps that come with every HGX, H200 NVL, Server Edition and NVL72 purchase, explained so your order ships on time.

4 min read

In this guide

  1. Why a GPU purchase comes with paperwork
  2. Which products are covered

Choosing a GPU server for LLM inference

A practical method for sizing an inference server from the model, the quantization and the concurrency you expect, with worked examples.

5 min read

In this guide

  1. Start with the model, not the server
  2. Step 1: how much GPU memory the weights need

HGX vs PCIe GPU servers

The real differences between an HGX node and a PCIe GPU server, and a simple rule for choosing between them.

5 min read

In this guide

  1. Two kinds of 8 GPU server
  2. The interconnect is the point

How to request a quotation

A short guide to getting a complete, accurate quotation, from the details we need to what the quotation contains.

4 min read

In this guide

  1. Why there are no prices on this site
  2. Three ways to ask

RTX PRO 6000 Blackwell editions explained

Workstation, Max-Q and Server Edition share one GPU and 96 GB of memory; this guide explains what differs and which to buy for a desk, a dense workstation or a rack.

5 min read

In this guide

  1. One GPU, three ways to cool it
  2. The three editions

Sizing CPU, memory and storage around GPUs

Six rules of thumb that keep the GPUs fed, with the arithmetic for an 8 GPU node, a PCIe inference server and a 16 node training pod.

5 min read

In this guide

  1. The GPUs are the expensive part, so feed them
  2. The six rules

Want this GPU in a different chassis?

We quote every class that carries it, from a single node to a pod. Tell us the workload and we propose the chassis.

Request a quotationView quote list