Teams at Ambarella, Inc.

Design, implement, and test imaging and video algorithms and DSP microcode for Ambarella SoCs. Develop firmware, research and implement DSP algorithms, optimize CPU and memory usage, create hardware-block test code, and debug assembly. The role requires understanding computer architecture, video standards, image processing, neural network frameworks, software development practices, and cross-functional teamwork.

Develop highly efficient deep learning and computer vision algorithms for edge applications, including self-driving cars and humanoid robots. Responsibilities include designing, training, and optimizing Transformer models, researching multimodal large language models and embodied AI, and building vision systems. The role also involves applying computer vision methods such as object detection, tracking, segmentation, and vision-language modeling, alongside robot control and sensor integration using ROS 2.

Develop software tools for future Ambarella embedded processors used in computer vision and intelligent automotive applications. Design and implement compiler language features, performance optimization tools, and related development capabilities. Collaborate with scientists, architects, and software and hardware teams to optimize chip performance. Work independently within a small R&D team.