You will work with software and hardware engineering groups to design the most energy-efficient AI accelerator and SoC

In this role, you’ll architect and develop runtime libraries that bridge the host system with Moffett’s computing device. You will interact with ML framework engineers and firmware/hardware engineers to identify opportunities and requirements.

In this role, you will deeply understand the specific needs in a field of ML application (e.g. classification, detection, segmentation, 3d reconstruction), and train/deploy ML models using toolchains developed by the company to achieve the best task precision and inference efficiency on Moffett’s computing platform

As the director of ML framework team, you will build, lead and manage a team of ML software engineers to develop the software toolchain and model zoos, and communicate with engineering team of our customers.

Implement performance and functional model for AI accelerator architecture exploration and functional verification

Drive Architecture/Software/Hardware co-design from system level requirement

You will work with the hardware and software teams to design an ASIC/FPGA system for CNN accelerate engine.

Understand and improve upon state-of-the-art algorithms in model compression and inference/training acceleration.

Develop inference engine that connects ML framework such as Tensorflow, Pytorch, or MxNet with customized platform (i.e. FPGA/ASIC).

Utilize existing frameworks (i.e. Tensorflow/Pytorch) and toolchains developed by the company to train models, compress models, and accelerate model inference.