This job involves the development of optimized DSP software libraries used in Signal Processing & Neural Networks domain. The job includes
Developing and Optimizing performance of signal processing firmware and neural network kernels using C level changes using C instrinsics.
Updating the DSP firmware for features, interfaces for different processor configurations.
Documenting the implementation for engineering details and for customers.
Experience with TensorFlow, TFLM, PyTorch, ONNX frameworks, TVM / GLOW compilers would be a plus.
The candidate should have a good experience in developing and optimizing DSP and NN applications in C and assembly language for a DSP or DSP-like processor, and should have a strong grasp of the issues surrounding software performance tuning.
Good technical communication skills are required. Candidate must be self-motivated and capable of working independently or as part of a team.
2 to 5 years’ relevant work experience in DSP firmware, and optimized neural network kernel implementations.
Educational qualification: BE/BS in CS or EE is required, MTech/MS is preferred. Education needs to be from a reputed institute – preference would be given to IITs, IISc, BITS Pilani.