This training is not provided as a standalone service but as an integral part of the comprehensive service packages offered to businesses, start-ups, and public administrations. It ensures that beneficiaries possess the foundational skills required to effectively utilize the computational resources provided for their AI initiatives.
The curriculum is designed to be accessible to users at all levels of prior experience, including those with no previous background in supercomputing. It bridges the gap between local AI development and large-scale infrastructure, enabling teams to scale their workflows autonomously and efficiently.
Practical example
The training modules are tailored based on the specific needs and HPC expertise of the beneficiary.
This service can cover, among other topics:
- Core HPC Operations: Introduction to HPC systems, user environment setup, access protocols, and best practices for job scheduling and submission.
- AI Workload Execution: Running AI workflows on supercomputers, including debugging, profiling, and managing complex AI workloads.
- Large-Scale Training: Multi-GPU and multi-node pre-training and fine-tuning techniques specifically designed for large-scale AI models.
- Optimization & Orchestration: Advanced optimization techniques and workflow orchestration to maximize computational efficiency.
- Trust & Governance: Deep dives into cybersecurity implications, risk management, and the regulatory/compliance obligations of AI on HPC.
