We need to do a large number of generate-prep and development builds as part of finishing up our new approach and automation for Pixel device support. Can anyone get us cloud computing credits? Otherwise, we need to start paying for multiple new Hetzner dedicated servers.

We do all production builds on 3 local GrapheneOS Foundation machines and each OS developer has at least one powerful local workstation. However, we need a lot more computing power than usual due to the way we’re adding back device support to AOSP requiring many clean builds.