PRIVATE AI SERVER

Frequently
asked
questions

• Who are Private AI Servers suitable for?

  Private AI servers are ideal for companies, institutions, developers, data scientists, researchers, and AI enthusiasts who want to implement and develop AI projects independently and autonomously, while also placing a high value on data security, data protection, and control over their sensitive information.

• What does “unmanaged” mean?

  We provide you with the necessary AI infrastructure through the Private AI Server, which includes a powerful server and an NVIDIA GPU. The management, maintenance, and acquisition of additional resources, such as an LLM, are entirely up to you. If you are interested in a managed service, our Managed LLM is the right product for you.

• What is a GPU?

  A GPU (Graphics Processing Unit) is a specialized processor originally developed for rendering graphics, but it has since become a key component in AI technology.

  Compared to a traditional CPU (Central Processing Unit), a GPU can handle millions of tasks in parallel, making it especially effective for complex computations and data-intensive applications like AI training, machine learning, and deep learning. On an AI server, the GPU significantly accelerates computational processes, enabling faster model processing and optimization.

• What is the difference between the NVIDIA A100 and H100 GPU?

  The NVIDIA A100 and H100 GPUs are state-of-the-art computing units specifically designed for AI applications. The main difference lies in the technological advancements of the H100 compared to the A100.

  The A100 is highly powerful, widely used and ideal for a variety of AI applications. The H100 offers higher computing performance, an optimized memory architecture, and new features for more complex AI processes and models. It can process larger datasets, more complex contexts, and parallel requests faster and more efficiently. Both GPUs are well-suited for the development of AI solutions.

• What is fine-tuning and what is it useful for?

  Fine-tuning is the process of adapting a pre-trained AI model to specific requirements or data. Instead of training a model from scratch, an existing open-source model that has already been trained on general data is used and “refined” with specific data relevant to the given application.

  This is particularly useful because it saves time and resources while increasing the model’s accuracy for the desired purpose. By fine-tuning, you can customize the model to better meet your individual needs – whether in natural language processing, image recognition, or other AI applications.

• What security standards does the Private AI Server meet?

  Deployment and operation take place in a dedicated Cloud environment (Private Cloud) to ensure maximum security for sensitive company data. By adhering to the strictest European and German data protection regulations, your valuable data is securely stored in our data center in Germany.

  Additionally, our data center is certified according to ISO/IEC 27001:2013 and meets the TIER 3 security standard for international data centers. Access to the server is provided via the Internet, protected by our firewalls.

• What does support cover?

  Cloudiax offers 24/7/365 support to ensure server availability on our end. Application consulting is not included in the support service.

• How is billing handled, and what is the contract term?

  Private AI Server A100 and H100: Billing is done quarterly in advance, with a minimum contract term of 12 months.

  Private AI Server for Fine-tuning: Billing is done on a weekly basis, with a minimum contract term of 1 week. The initial setup of the Private AI server for fine-tuning takes place on the first day of usage.

  All AI servers are billed at a flat rate and not based on usage at the stated price.