ASIC Design Engineer (Deep Learning)

NovuMind is a startup co-located in Silicon Valley and Beijing. Our mission is to eliminate the existing barriers of artificial intelligence deployment, so that companies of all sizes, in all industry sectors, can unleash the full power of AI. Centered around our AI chip, our full stack solutions of deep learning enable AI in products and services from embedded to cloud. We are looking for an ASIC Design Engineer (Deep Learning) in our Santa Clara, CA office.

Duties & Responsibilities

• Responsible for neural network engine design. Design computer arithmetic and data path circuits for convolution neural networks.
• Perform micro architecture design, write specification, and implementation RTL. Perform linting checks, module level simulation, synthesis and debug timing.
• Perform complexity trade off analysis, and computer arithmetic optimization
• Work with scientists on algorithm improvement. Work with design verification and system teams, provide inputs on test plan and firmware implementation.
• Perform performance, power, area trade off analysis to achieve optimal logic efficiency.

Essential Qualifications

• Master/PhD degrees required in Computer Science, Computer Engineering, Electrical Engineering
• Verilog/System Verilog language
• 2-12 years of industry experience. in one or more fields below:
• Digital Signal Processing (FIR filters, FFT transforms, encoder/decoders etc.) and Computer Arithmetic Circuits
• Artificial Neural Networks, Deep Learning
• Computer Architecture, CPU or GPU
• Familiar with Micro architecture Design, Verilog implementation, Linting, performance vs. power trade offs.
• Highly motivated and self-disciplined. Strong verbal and oral communications skills. Strong team work skills, and ready to take technical leadership.

Preferred Requirements

• Understanding of basic synthesis flow, Statistic Timing Analysis and timing debugging.
• Scripting skills (Perl, Makefile, Python, TCL)
• FPGA prototyping experience