Lightelligence develops optical computing platforms for high-performance computing tasks. By processing information with light, our custom chips potentially offer orders of magnitude improvement in latency and low power consumption compared to traditional electronic architectures. If you would like to build computational machines that are solving problems in ways that have never been done, we are looking for software engineers with strong problem-solving skills to help us build an SDK for our optical hardware accelerators. This SDK converts a trained neural network into a binary we can run on our custom hardware. The SDK imports neural networks from popular frameworks (e.g., TensorFlow, PyTorch) and runs a series of transformations on those neural networks. These transformations include things like quantizing certain parts of the graph and partitioning subgraphs that can run on our device. The SDK also includes a compiler and runtime library to run the transformed neural network on both a simulator and our actual hardware. What you will do at Lightelligence: Create a customer portal to help keep track of bugs in the SDK Write documentation and literature for customers to learn how to use the SDK Debug and refactor parts of the SDK based on customer feedback Work with electrical engineers to update compiler and runtime libraries Create profiling and debugging tools for both internal and customer use Basic Qualifications: Bachelor’s degree in computer science or related fields 2+ years of experience writing industry standard code Proficient in at least one object-oriented programming language (i.e., Python, Java, C++) Familiarity with Python and C++ Experience with software testing and building robust software systems Strong debugging and problem-solving skills
某知名AI光子公司某知名AI光子公司某知名AI光子公司
职位顾问