How many people these days have the opportunity to create a new processor ISA? At Plumerai, you will have that opportunity!
We create embedded Deep Learning hardware accelerators, combining a new generation of neural networks with our novel computing architecture.
We are looking for a low-level programmer with experience in bare-metal programming, compilers and parallel processing. Having FPGA experience will be a huge plus, as well as experience with neural networks and frameworks (Tensorflow, Torch, MXnet, etc).
You will be responsible for the development of an interpreter to convert a high-level network model into a sequence of machine-readable code for our proprietary massive parallel architecture. The architecture is in active development, so you will have an influence on what commands and features will be added to it.
What you need:
- Knowledge of how compilers work.
- Experience building tools like GCC rebuild, lex, yacc or similar.
- Excellent understanding of processor architectures and low-level programming (assembly/opcode level).
- Practical experience with low-level parallel programming (CUDA, OpenMP).
- Good mathematical skills.
- Knowledge of computational and memory complexity.
Highly desirable and nice to have:
- Experience with queuing theory and dynamic programming.
- Experience in writing Linux kernel modules and/or device drivers.
- Experience with hardware emulation software and techniques (QEMU).
- FPGA experience.
- Knowledge of neural networks and modern deep learning frameworks.
What you will get:
- Paid travel to top research and developer conferences.
- Competitive base salary.
- Choose your own laptop.
- 25 days of paid vacation time in addition to bank holidays.
- Flexible working hours.
- Company stock options package.
- Pension contribution.