ABOUT THE PROJECT:
Our client is an innovative startup from Silicon Valley which is operating in stealth-mode building next-generation semiconductor chips targeted for data centers and future Edge Computing deployment. You will have a unique opportunity to participate in one of the most innovative ML\Hardware projects to build a chip for AI acceleration
You will have a chance:
-to work in the international team for the next generation of AI acceleration chips
— deep dive into AI solutions Grow your experience in writing highly optimized code
MAIN TASKS AND RESPONSIBILITIES
The team is working on a new processor-accelerator for neural networks in 2 directions: implementation neural network operations on hardware with custom Tensorflow development (C / C ++ / Python), neural network graphs optimization through Tensorflow Graph Transform Tool and test environment creation (Python).
The engineers in both directions work closely, so the ability to read/write code in C/C ++ will be a plus since the achievements of low-level implementations should be used in graph transforms workflow.
The current position implies work on graph transformation and Python and Tensorflow knowledge is a must.
You will be responsible for delivering highly optimized, documented, clean and maintainable Python code and suggesting technical and functional improvements to add value to the product.
At this position, you will be able to understand in detail well-known neural network architectures, learn how to optimize their graphs, and also take part in optimizing completely new, custom architectures that solve quite interesting problems.
MUST HAVE (skills\knowledge\experience):
5+ years in software development or masters in Computer Science (Systems, Arch, Compilers, AI, ML)
Expert Python programming experience
Expert level experience in Tensorflow framework development
Be familiar with Tensorflow graph transform tool
Experience in Tensorboard and neural networks profiling
Neural networks optimization (quantization, fusing, folding)
Experience in deep learning models for image analysis (CNN)
Experience in language models (RNN)
WOULD BE A PLUS (skills\knowledge\experience):
Familiarity with other deep learning frameworks and libraries: PyTorch, ONNX, Caffe, Scikit, etc
Familiarity with fixed-point arithmetic & quantization
C++ programming experience
— Friendly and highly professional teams
— Flexible working hours with no overtimes
— Regular performance reviews
— Comfortable office facilities (kitchens, shower, gym, sports activities, yoga, lounge zone, coffee points etc.)
— Fully paid English classes (twice per week) with own English teachers
— 20 working days of annual paid vacation
— Christmas holidays
— Paid sick-leaves
— Premium Medical & Life insurance
— Incentives (marriage, childbirth)
— Training activities & bonuses for internal training delivery
— Corporate events