GPU
Speed up your projects with our high-performance GPU instances
Deploy high-performance NVIDIA® instances in seconds. Ideal for machine learning, 3D rendering, and large-scale processing. Hosted in Switzerland to ensure data sovereignty and low latency.
Dedicated GPU power
Direct access to GPUs. Zero latency due to virtualisation; raw performance guaranteed.
Sustainable performance
GPUs hosted in data centers powered by 100% renewable energy. High performance with a reduced carbon footprint.
Easy cloud Integration
OpenStack deployment. Automation via Terraform and API for easy management.
Accelerate your innovation with dedicated machines
Artificial Intelligence & machine learning
Ultra-fast training of neural networks (LLMs, CNNs). Real-time, low-latency inference – essential for vision and NLP in production.
3D rendering, video encoding and streaming
Massively accelerated photorealistic rendering and VFX. High-density video transcoding without bottlenecks for streaming.
High-performance computing and big data
Optimised complex scientific simulations (financial, climate). Processing and visualisation of petabytes of data with maximum efficiency.
Your ideal GPU instance
We offer a strategic selection of NVIDIA instances to meet all your needs, from the development phase through to intensive production.
Inference
NVIDIA B300
Push the boundaries of generative AI with the B300 Blackwell Ultra GPU, ideal for inference of massive LLMs and reasoning.
288 GB of HBM3e VRAM
~15 PFLOPS of FP4 computation (dense)
8 TB/s of memory bandwidth
H100 GPU instances
Train and deploy your LLMs on a large scale with the H100 Hopper GPU, ideal for training massive models, low-latency inference and intensive HPC.
80 GB of HBM3 VRAM
~3,958 TFLOPS of FP8 computation
3.35 TB/s of memory bandwidth
NVIDIA L40S
Take your generative AI projects to the next level with the L40S GPU, ideal for training large language models (LLMs), high-fidelity 3D rendering and extreme graphics performance.
48 GB of GDDR6 VRAM
~1,466 TOPS of inference (INT8)
864 GB/s of memory bandwidth
NVIDIA A100
Harness the power of the A100 GPU for intensive deep learning and HPC, delivering record-breaking bandwidth for massive datasets.
40 GB of HBM2e VRAM
~1,248 TOPS of inference (INT8)
2,039 GB/s of memory bandwidth
NVIDIA T4 GPU
Accelerate your AI projects with versatile T4 GPUs, ideal for inference, multimedia processing, and lightweight deep learning.
16 GB of GDDR6 VRAM
~130 TOPS of inference (INT8)
320 GB/s of memory bandwidth
NVIDIA L4
Optimise your AI workloads with high-performance L4 GPUs, ideal for advanced inference, video processing and intermediate deep learning models.
24 GB of GDDR6 VRAM
~260 TOPS of inference (INT8)
300 GB/s of memory bandwidth
NVIDIA A2
Deploy your lightweight AI models easily using energy-efficient A2 GPUs, ideal for basic inference and essential multimedia tasks.
16 GB of GDDR6 VRAM
~170 TOPS of inference (INT8)
200 GB/s of memory bandwidth
Perfectly integrated into Infomaniak's Public Cloud
Discover a first-class cloud experience with seamlessly integrated services designed to work together.
Cloud instances
Linux instances
GPU instances
Optimised instances
Storage
Block storage
Object storage
Snapshots & backups
Network
DNS as a Service
Load balancers
DDoS protection
The greenest AI infrastructure worldwide
We invest heavily in R&D to reduce the environmental impact of artificial intelligence.
100% local renewable energy
In addition to purchasing only locally generated renewable electricity, Infomaniak is building its own solar power plants using European panels with a low carbon footprint. The goal is to cover 50% of our energy needs through self-generation by 2030.
100% energy recovered to heat households
The renewable energy that powers our AI services is fully recovered to heat 6,000 homes in winter and provide the equivalent of 20,000 showers per day in summer. The GPUs are cooled without dedicated air-conditioning and use only a small quantity of water in the event of a heatwave, thanks to the adiabatic principle.
Certified B Corp
With an average PUE of less than 1.1, a WUE of 0.0145 and carbon offsetting totalling 200% of our emissions, we are building a sovereign and sustainable European cloud, designed for future generations. This commitment is based on strict, internationally recognised standards.
A sovereign cloud ecosystem that guarantees control over your data
Infomaniak's cloud solutions provide companies and public authorities with the guarantee that their data remains under Swiss jurisdiction, in full compliance with the RGPD.
100% sovereign Swiss cloud
Infomaniak operates its own data centers in Switzerland and develops its solutions in-house, in Geneva and Winterthur.
Respect for privacy and data
Infomaniak is owned by its employees. Your data is processed by software developed in Switzerland that respects RGPD, nFADP and privacy.
Data security and protection
Infomaniak develops its services and infrastructures from A to Z to guarantee data security and confidentiality.
Try it for free now
Create a bucket, connect your tools and start using sovereign, scalable and transparent cloud storage.
Get started for free
New project?
Talk to a cloud expert
Our engineers have helped hundreds of customers to scale and start their cloud infrastructure intelligently.