At the Pioneer Centre AI, we have prioritized minimizing day-to-day friction for our researchers. The centre is investing heavily in two compute installations set for continual expansion. Both installations will be tailor-made for AI workloads, allowing our researchers to take their code from conception to being ready for final training at larger facilities in a nurturing environment with room for experimenting, making mistakes and growing. One of these facilities will be for GDPR-compliant computing, allowing the centre’s researchers access to highly restricted data. Facilities will be available to all researchers affiliated with the Pioneer Centre. Researchers will also have the option to bring their own compute budgets for having their own priority resources within the Pioneer Centre for AI’s facilities.
To help users and staff share and disseminate information across universities, the centre has an internal documentation system featuring information such as computing workflows, where and how to get compute hours at larger facilities, how to use the compute facilities, as well as educational material about technical topics. The Pioneer Centre for AI hosts intro talks from experts on topics such as GPUs, HPC systems and graph-based computing systems, with active participation and discussion being particularly welcome. Frequently, researchers from universities across the globe will give talks about their research.
To make sure our researchers have potential obstructions removed from reaching their goals, the Pioneer Centre for AI provides dedicated computing personnel who help with:
- Sparring about code
- Advice about where and how to run code
- Assistance in finding compute hours at bigger facilities
- Suggestions for how to solve errors
- Custom low-level code written
- Advice on how to spend computing budget