Engineering Director – Neural Network Compiler

At the Tensilica IP division of Cadence we design customizable Xtensa cores as building blocks for highly optimized domain-specific embedded platforms. With our energy-efficient DSPs and the industry leading DNA processor family, we look to dominate the embedded applications market for deep learning inference engines.

As Tensilica’s Engineering Director of the Neural Network Compiler team, you will be driving the direction and managing the implementation of our entire AI software stack. You will play a key role in furthering our success in the fast-growing machine learning market segment and have an opportunity to impact many customers who use our technology to advance their machine learning products.

Responsibilities

Manage and grow a world-class team of software engineers working on Tensilica’s Neural Network Compiler. Define project scopes and milestones and drive the on-time execution.
Provide technical leadership in defining forward-looking projects and novel software concepts to significantly impact Tensilica’s machine learning product portfolio and roadmap.
Work closely with Tensilica’s architecture and hardware implementation teams to guide the specification and development of future neural network processing engines.
Cooperate with Tensilica’s core SDK team to continuously improve mapping of machine learning algorithms onto our customizable DSP cores and DNA neural network processors.
Collaborate with Tensilica’s application engineering and marketing teams to understand the needs of current and potential future customers and define software solutions that will resolve their real-world machine learning problems.
Requirements

PhD or Master’s degree in Computer Science or Engineering with 12+ years of experience in a field of compiler development or machine learning applications.
Expertise in compiler construction in a production environment with demonstrated ability to deliver high-quality solutions for customer needs.
Proven track record of building and managing top-notch software development teams.
Algorithmic depth in the field of neural network processing with expertise in compiler mapping of machine learning algorithms to various architectures.
Solid understanding of computer architecture concepts and strong background in optimizing code for non-conventional processors (e.g., VLIW DSPs or specialized HW accelerators).
Strong software development skills in C++ and Python.
Good knowledge of one or more machine learning development projects (TensorFlow, PyTorch, ONNX, TVM).