Myrtle is looking for multiple skilled Haskell developers to join our growing development team writing a compiler for generating highly efficient hardware designs implementing machine learning solutions.
We use Haskell for every aspect of our work and developers will get the chance to be involved in every part of the project. As an example, some things we’ve worked on recently are a tiny CPU to implement a custom network protocol to communicate between a PC and an FPGA and a Haskell implementation of part of the Caffe deep learning framework.
Specific tasks you could be working on soon include: implementing fixed point neural network evaluation support in our compiler, supporting AWS F1 instances as a backend, investigating and implementing exciting techniques for neural network evaluation such as the Winograd small convolution algorithm and realising full hardware implementations of existing neural networks.
A familiarity and interest with hardware design would be viewed positively but is not essential. Applicants should be familiar with Unix based systems; all our development is done on Ubuntu or NixOS.
Applicants should send a resume and covering email, stating your suitability for this role, to firstname.lastname@example.org. Please include any links to blog posts, repositories or any other media which demonstrates your skill set.
Roles will be based at our offices in Cambridge, UK.
Closing date for applications 5pm, 15th March 2017
Myrtle is a software company based in Cambridge UK with a long history of working at the forefront of compiler technology. We are best known for having helped produce computer generated images for over twenty major Hollywood movies including the Transformers franchise. Myrtle’s clients have included NYSE and NASDAQ listed companies in LA, Vancouver and London.