Senior Deep Learning Hardware Architect

Main responsibilities
• Research and develop high-level architecture and micro-architecture for neural
processor hardware.
• To optimize execution of neural network models on neural processor hardware.
• Expertise in deep learning-related algorithms, such as Convolutional Neural
Networks, Long Short-Term Memory, classification/detection, and their
applications.
• Familiarity with computer vision algorithms such as object detection, tracking, and
recognition.
• Understand the hardware implications of the aforementioned algorithms in terms of
performance and power.
• Implement neural processor behavior, functional, and cycle-accurate models (e.g.
SystemC or C++)
• Help implement NPU synthesizable code, simulate, debug, and optimize.
• Benchmark models, simulations, and prototypes.

Minimum Qualification
Masters, PhD, or equivalent experience in Computer Engineering, Electrical
Engineering, or related field.
• Knowledge/expertise in hardware architecture, computer architecture, and GPU
architecture.
• Proven C++ and Python coding skills.
• Familiarity with neural networks and machine learning concepts, neural processor
architectures, network on chips.
• Desirable – Experience with using and modifying TensorFlow, Caffe, & PyTorch
code and/or other neural network development frameworks is a plus.
• Desirable – Knowledge of image processing, camera pipeline is a plus.
Additionally, we look for the following universal qualities in all candidates:
• Resourceful Achiever: self-driven and proactive. You apply logic and reason to
effectively solve problems and manage risks.
• Avid Learner: you eagerly take on new challenges and seek out opportunities to
grow and stretch.
• Passionate Owner: you are energized by your work, taking ownership, and
delivering results without ego.
• Committed Collaborator: with a positive attitude and commitment to get to the best
result, you welcome ideas from others and drive processes forward in an inclusive
manner