Cloud GPU – from €92.59 per month · available immediately

Rent a GPU server: NVIDIA Power for AI

Dedicated NVIDIA RTX A4000 and RTX 6000 Ada GPUs in the centron Cloud – maximum GPU performance, certified security and 100% data sovereignty, Made in Germany.

  • Dedicated NVIDIA GPUs – no shared resources
  • Ready to go in 2 minutes – no contract commitment or minimum term
  • Linux or Windows Server – the system that suits your project
  • Usage-based billing at clearly predictable prices
ccloud³ Console · New GPU instance
eu-de · Hallstadt Data Centre
GPU
RTX 6000 Ada
VRAM
48 GB
Billing
per hour
GPU NVIDIA RTX 6000 Ada · dedicated
● Selected
OS Ubuntu 24.04 LTS · CUDA-ready
● Ready
SSD NVMe storage · 200 GB
● Ready
BK cBacks · automated backups
● Enabled
GPU instance deployedRTX 6000 Ada · 96 s
High performance from Germany

Full GPU power for AI, rendering and HPC

Train AI models, speed up rendering and scale computationally intensive workloads – powered by NVIDIA and operated on German infrastructure that complies with the GDPR.

Dedicated GPU performance

No shared resources: You get the full power of your GPU for AI training, inference, 3D rendering and HPC workloads – reproducible every time.

Available in a matter of minutes

Start, pause, scale up or down GPU instances – in line with your project phase. This allows you to cope with peaks in demand and keep budgets predictable.

Full cost control from €92.59 per month

No minimum contract term, usage-based and with no hidden costs or minimum contract terms. There is absolutely no need for significant upfront investment in your own hardware.

Data sovereignty included

Your data is stored in an certified to ISO 27001 on the basis of IT-Grundschutz, high-security data centre in Germany – GDPR-compliant and with an unrestricted BSI C5:2020 Type 1 certificate.

From self-service to fully managed

Running GPU workloads without owning any hardware

Whether it’s AI inference, machine learning, big data, video transcoding or simulation: centron provision scalable GPU instances in a matter of minutes – with managed packages for patching, observability, backup/recovery and contingency procedures available on request, to take the pressure off your team.

from €92.59 per monthNVIDIA RTX A4000 to A100 · dedicated
Calculate GPU costs in the price calculator
  • Dedicated NVIDIA GPUs – no shared resources
  • Ready to go in 2 minutes – no contract commitment or minimum term
  • Linux or Windows Server – the system that suits your project
  • Usage-based billing at clearly predictable prices
An overview of GPU instances

Choose the GPU that suits your use case

From energy-efficient inference to training large models – all GPU classes run in dedicated mode with no minimum runtime.

NVIDIA RTX A4000
16 GB VRAM
Ab
92,59 € / Monat
The all-rounder
  • CAD & 3D Modelling
  • Video editing & rendering
  • AI-powered image processing
NVIDIA Quadro RTX 6000
24 GB VRAM
Ab
170,83 € / Monat
The powerhouse for simulation and AI
  • 3D Animation & VFX
  • VR & Simulation
  • Workstation AI Training
NVIDIA A100
40 / 80 GB VRAM
Ab
489,47 € / Monat
The AI & HPC Monster
  • AI model training
  • High-Performance Computing (HPC)
  • Big Data & Analytics
NVIDIA RTX 6000 Ada
48 GB VRAM
Ab
858,19 € / Monat
Ray Tracing & AI Power
  • Real-time ray tracing & rendering
  • Deep Learning & AI
  • Research & Simulation
Renting GPU servers – an overview of prices

How much does a GPU server cost at Centron?

A GPU server with centron starts at €92.59 per month (NVIDIA RTX A4000, billed by the hour). All GPUs run on a dedicated basis – you do not share the processing power with anyone – in certified to ISO 27001 on the basis of IT-Grundschutz data centres in Germany, with no minimum contract term and a €200 starting credit for new accounts. They can also be used as AI servers for training and inference.

A comparison of GPU models
GPU modelVRAMIdeal forPrice per month
NVIDIA RTX A400016 GBAI inference, CAD, rendering, image processingfrom €92.59
NVIDIA Quadro RTX 600024 GB3D animation, VR, simulation, workstation AIfrom €170.83
NVIDIA A10040 / 80 GBAI Training, HPC, Big Data & Analyticsfrom €489.47
NVIDIA RTX 6000 Ada48 GBReal-time ray tracing, deep learning, researchfrom €858.19
Success stories

How our customers manage their IT

Echte Praxisbeispiele von Kunden, die ihre IT-Infrastruktur mit centron neu aufgestellt haben – von Skalierung unter hoher Last über sichere Datenverarbeitung bis zur Kostenoptimierung.

Take the next step

Register now & get a €200 starting balance

Instant high performance – without having to buy or maintain your own hardware
Complete flexibility – Start, pause and scale instances as required
Predictable budgets – usage-based billing with no fixed costs
€200 trial credit – Get started straight away with your GPU use case
Cloud GPU FAQ

Everything you need to know about centron Cloud GPUs

Which GPUs are available from Centron?

Currently available are dedicated NVIDIA RTX A4000, Quadro RTX 6000, A100 (40/80 GB) and RTX 6000 Ada GPUs – ready to go straight away for AI inference, machine learning, big data, rendering and simulation. You’ll always get the full performance of your GPU, without any shared resources.

How soon can I get started?

In just 2 minutes: create an account for free, configure a GPU instance, and get started – with no contract and no minimum term. New accounts receive a €200 starting credit to try out the service.

How much does a cloud GPU cost?

Prices start from €92.59 per month (NVIDIA RTX A4000). Billing is usage-based and transparent – with no hidden costs. You can calculate the cost of your configuration using the price calculator.

Where is my data stored?

At the certified to ISO 27001 on the basis of IT-Grundschutz high-security data centre in Germany. The service is GDPR-compliant; ccloud³ / Managed Cloud also holds an unrestricted BSI C5:2020 Type 1 certificate – ensuring 100% data sovereignty, Made in Germany. Find out more: Data sovereignty at centron.

Can Centron take over the operation?

Yes. From self-service to fully managed: managed packages cover patching, observability (logs, metrics, traces), backup and recovery, and contingency procedures. With cProtect, replication takes place every 15 minutes and failover can be completed in a matter of seconds.
Read more

AI workloads and vendor comparison

Get started for free

Sign up and receive €200 credit at centron within your first 60 days.

This promotional offer applies to new accounts only. Available exclusively to businesses.

Jetzt loslegen Sales kontaktieren