Custom Deep Learning Workstations from 3XS

AI Development Workstations by 3XS Systems

As an NVIDIA Elite Partner, we pioneered the concept of the AI workstation in 2016 to complement the original NVIDIA DGX.

3XS AI workstations make starting your AI journey easy. Powered by NVIDIA GPUs, they provide data scientists with a cost-effective platform for developing AI models, prior to scaling up to datacentre-grade training hardware which is also available from Scan.

Designed for office environments, 3XS AI workstations are fine-tuned by our hardware engineers and data scientists for performance, reliability and low-noise. We provide a custom software stack that means you spend more time on your code development and less time configuring drivers, libraries and APIs.

Enquire Now

NVIDIA GPU Options

There are two types of GPU available, browse the different options below.

NVIDIA RTX PRO

NVIDIA RTX PRO GPU-accelerators provide the highest performance and largest onboard memory for AI workstations. The latest generation of RTX PRO cards are based on the Blackwell architecture, which added native support for FP4 precision, and select models support Multi-Instance GPU (MIG), enabling multiple users to work on the system at the same time.

NVIDIA GeForce RTX

NVIDIA GeForce RTX GPU-accelerators provide mid-range performance and smaller onboard memory at a lower price point than RTX PRO cards. The latest generation of GeForce RTX cards are based on the Blackwell architecture, which added native support for FP4 precision. GeForce RTX cards do not support Multi-Instance GPU (MIG).

	RTX PRO 6000 Blackwell	RTX PRO 6000 Blackwell Max-Q	RTX PRO 5000 Blackwell	RTX PRO 4500 Blackwell	RTX PRO 4000 Blackwell	GeForce RTX 5090	GeForce RTX 5080
Architecture	Blackwell	Blackwell	Blackwell	Blackwell	Blackwell	Blackwell	Blackwell
CUDA Cores	24,064	24,064	12,800	10,496	8960	21,760	10,752
Tensor Cores	752 5th gen	752 5th gen	400 5th gen	328 5th gen	280 5th gen	680 5th gen	336 5th gen
RT Cores	188 4th gen	188 4th gen	100 4th gen	82 4th gen	70 4th gen	170 4th gen	84 4th gen
Memory	96GB GDDR7	96GB GDDR7	48GB GDDR7	32GB GDDR7	24GB GDDR7	32GB GDDR7	16GB GDDR7
ECC Memory	✔	✔	✔	✔	✔	✖	✖
Memory Controller	512-bit	512-bit	384-bit	256-bit	192-bit	512-bit	256-bit
NVLink	✖	✖	✖	✖	✖	✖	✖
MIG	4 instances	4 instances	2 instances	✖	✖	✖	✖
TDP	600W	300W	300W	200W	140W	575W	360W

Learn more about RTX PRO and GeForce RTX graphics cards in our buyers guides.

GPU Configurations

3XS AI workstations can be configured with two air-cooled GPUs, or up to eight GPUs when watercooled. Adding more GPUs not only improves processing speed, it also enables you to run larger models with more parameters. Use the table to identify what type of configuration is required for your projects.

GPU	RTX PRO 6000 Blackwell			RTX PRO 5000 Blackwell 72GB		RTX PRO 5000 Blackwell 48GB		RTX PRO 4500 Blackwell		RTX PRO 4000 Blackwell		GeForce RTX 5090			GeForce RTX 5080
Number of GPUs	8	2	1	2	1	2	1	2	1	2	1	8	2	1	1
Cooling	Water	Air	Air	Air	Air	Air	Air	Air	Air	Air	Air	Water	Air	Air	Air
AI performance (FP4 TOPS)*	32,000	8,000	4,000	4,128	2,064	4,128	2,064	3,200	1,600	2,356	1,178	26,816	6,704	3,352	1,801
GPU memory*	768GB	192GB	96GB	144GB	72GB	96GB	48GB	64GB	32GB	48GB	24GB	256GB	64GB	32GB	16GB
Billions of AI model parameters (FP4)*	1,224	307	153	230	115	153	77	102	51	77	38	408	102	51	26

* combined figure for all GPUs

3XS AI workstations are one of several types of desktop device suitable for AI model development. Read our AI workstation buyers guide for information on how they compare.

Maximise Your AI Workstation With Production Ready Software

NVIDIA AI Enterprise, available to purchase as an upgrade with RTX PRO GPUs, is a suite of frameworks and libraries, including NVIDIA NIM and NeMo microservices, that accelerate the development and deployment of AI projects. Leveraging pre-configured frameworks removes many of the manual tasks and complexity associated with software development, as each framework is tried, tested and optimised by NVIDIA. The less time spent developing, the greater the ROI on your AI hardware and data science investments.

Rather than trying to assemble thousands of co-dependent libraries and APIs from different authors when building your own AI applications, NVIDIA AI Enterprise removes this pain point by providing a full agentic AI software stack.

Find Out More

NVIDIA AI Enterprise software stack diagram

Develop on your choice of OS

3XS AI workstations are provided with a custom software stack to speed up your AI model development. Powered by NVIDIA GPUs, 3XS AI workstations are the perfect tool for unlocking the comprehensive catalogue of optimised software tools for AI and HPC provided in the NVIDIA GPU Cloud (NGC).

Linux

Your system can be configured with the latest Ubuntu operating system, plus Docker for creating, sharing and running individual containers on a single node and TensorFlow, a popular framework for building and deploying AI applications.

Windows

Your system can be configured with the latest Microsoft Windows 11 operating system, using Windows Subsystem for Linux (WSL) to build and deploy your AI models.

3XS AI Workstations Benefits

Connected AI compute nodes linked by a glowing network path to illustrate scalable AI infrastructure.

Developing AI models

3XS AI workstations are specifically designed for the computationally-intensive task of developing AI models, whether you’re building an LLM, agentic AI or physical AI with robots.

The design combines powerful NVIDIA GPUs alongside the optimum CPU, RAM and storage configuration to ensure the fastest AI performance and compatibility with the NVIDIA software ecosystem.

High-performance workstation computer on a modern office desk beside a monitor displaying code.

Easy to deploy

Unlike most AI hardware, 3XS AI workstations are designed for office environments, using state-of-the-art air- and watercooling to keep noise to a minimum and industry-standard Ethernet networking.

This means they can be deployed in minutes, without requiring specialist knowledge and infrastructure, or approval from centralised IT departments.

Workstation computer with a digital processing interface and animated data streams representing AI workloads.

Keep your data local

3XS AI workstations process data locally, so are ideal for when you're unable to use cloud or centralised compute resources for your AI model development because of your organisation's data compliance policy.

Fast in-box processing means your data stays private and secure, even when multiple users are sharing access using MIG with RTX PRO GPUs.

Upgrade path from an AI workstation to an enterprise AI server and humanoid robot connected by arrows.

Purchase with confidence

As an NVIDIA Elite Partner, we pioneered the concept of the AI workstation in 2016 to complement the original NVIDIA DGX. Since then, we have continued to innovate, adding watercooling configurations with up to eight GPUs for incredible performance.

When you're ready to scale-out your new AI model we also sell NVIDIA DGX datacentre servers plus a wide range of robots for physical AI projects.

Start your AI journey

3XS AI workstations are available as preconfigured models or can be customised to your requirements. Alternatively, you can develop AI models on any device via virtualised NVIDIA GPUs hosted on our Scan Cloud platform.

Pre-Configured 3XS AI Workstations

Buy Now

Configure Your 3XS AI Workstation