Raspberry Pi foundation
supports our educational initiative since 2017 to teach students and researchers
how to develop sustainable research software, share artifacts, and
automatically optimize and test (autotuning, crowd-tuning and crowd-fuzzing)
realistic workloads in terms of speed, size, energy usage, accuracy and costs
across diverse software and hardware stack
using CK workflow framework and
open optimization repository.
We regularly help various international projects (MILEPOST,
and assist scientists in crowdsourcing and reproducing experiments,
and developing customizable and sustainable research software
powered by CK
which can now survive in a Cambrian AI/SW/HW chaos
or when leading researchers leave!