Become a sponsor to Subdivision AI
The best way to solve a big problem is to divide into small parts. An efficient distributed system is not a monolith, but a network of interacting components. As we survey the landscape of efficient computation, subdivisions abound.
At Subdivision AI, we identify computational challenges that advance the state of the art in machine learning. We share solutions so that the future of AI is accessible to all. We maintain our independence to ensure the diversity of the ecosystem.
Meet the team
-
Andrew Lavin andravinFounder. Specialist in efficient software and algorithms for deep learning.
$1,000 a month
SelectHelp fund cloud computing costs that contribute to research and development of efficient convolutional neural network models.
The best models will be shared through reference implementations and freely available tech reports.
$5,000 a month
SelectHelp fund development of efficient open source software for training and inference.