With our GPU-as-a-Service solution you can do anything. But to get you started faster, we offer you pre-built Docker images.
With these images you can for example immediately start with your Jupyter application. CUDA? Already installed. Python? Already installed. Pytorch? Already installed. Jupyter Lab? Already installed.
Your machine learning or deep learning application has never been deployed so easily and fast.
Modern applications require a comprehensive foundation. Even more so when it comes to frameworks and dependencies. With our pre-built ExaMesh images we offer you the perfect basis to start your own application.
See the following table for the current images we offer and with which components they are built:
Our stack is based on Docker. Therefore, it is possible that any Docker image can be deployed on our decentralized cloud.
You can use your own or any other public or private image. Whether it’s pulled from Docker Hub, Amazon ECR, GitHub, your private or any other registry powered by the open source Docker registry server.
You don´t even know how Docker works or you´re just familiar with the basics? No problem.
Due to our easy setup everybody can deploy an HPC Instance, e.g. with an NVIDIA GPU, with just a few clicks. Even if you don´t know what a compose file or YAML is.
Choose an ExaMesh image and just deploy it. No need to update the automatically generated Docker compose or anything else.
Our decentralized and 100% green infrastructure uses no additional virtualization other than Docker. No virtual hardware like vCPUs. Just dedicated CPU cores (each with two threads), RAM, SSD and of course GPUs.
This allows us to achieve maximum performance, agility and simplicity. Exactly what other large providers can no longer provide.
ExaMesh Instances: Cheap, simple, super fast and environmentally friendly.
Already in 2019, we were the first provider to run HPC Instances with 100% green energy. We don’t just use co2 certificates to do this. All of our servers run inside green power plants.
Here you can see how we install the V100 Instances in one of our wind turbines.
Whenever and no matter what the request. We are there for you.
Accounts can be used by multiple members of your team and manage any number of Instances and SSH keys.
Management of each Instance is always possible via SSH and your command line.
Nobody knows the exact location of your Instance. Our infrastructure is decentralized.
