This month marks the introduction of a completely revised system for managing access to GPU compute instances within the Data Science extension.

We are launching a dedicated credits management interface that enables organizers to precisely allocate compute credits to both fellow organizers and learners. This system runs across the three levels of administrative hierarchy within an organization - platform, activity, and individual - allowing credit allocation to cascade from one level to the next.

‍

‍

For data challenges or hackathons, as an alternative to the personal allocation, compute credits can be allocated at the group level, ensuring equitable access to GPU instances and fair competition among all participating groups.