Roche Sequencing is developing ground-breaking next-generation sequencing products that allow scientists/clinicians powerful new avenues to investigate DNA, the blueprint of any lifeform. We develop a proprietary system that generates data at tremendous rates, a big challenge for existing technologies for data processing. As we push the boundaries in DNA sequencing technology to make genetic information universally available through massively parallel DNA sequencing, we need you! The Opportunity: We are looking for a GPU and Deep Learning Software Engineer to develop GPU-accelerated software utilizing our proprietary models targeted for real-time analysis on our DNA sequencing instrument. The ideal candidate should have significant experience with GPU programming, an in-depth understanding of the architecture and performance characteristics of modern GPUs, together with a good understanding of neural network modeling and their implementation at a low level. We are also looking for candidates with the drive and capability to expand their knowledge as needed in the above areas. Responsibilities include: Develop GPU-accelerated software targeted for real-time analysis on a DNA sequencing instrument. Implement neural network algorithms on GPU, optimized for our unique characteristics of very high throughput and model types. Be proficient in C++ and CUDA, and have familiarity in python for modeling-related tasks. Self-motivation to individually research and design solutions for complex technical problems. Author dependable, readable, maintainable and well-structured code, and encourages best practices with the team. When you join us, you will have meaningful work developing an innovative technology that saves and improves lives around the world by making a huge difference in disease diagnostics! Our team is collaborative and fast-paced. And, we offer a competitive salary with a rich benefits package. Who you are: BS in Computer Science, Computer Engineering, or related field plus 5 years of direct related experience; MS/PhD is a plus. Or, a combination of education and experience to perform at this level. Object-oriented programming in C++ and experience with GPU-accelerated C/C++ libraries such as CUDA, cuDNN, or Thrust. Advanced experience in systems programming, specifically writing, debugging, and optimizing parallel (CPU and GPU) Linux applications. Good understanding of machine learning, particularly neural networks. Experience with a deep learning framework, such as PyTorch, JAX or TensorFlow, and how models are represented and transferred for efficient inference on GPUs. Experience training models is an advantage. Basic experience with Python. Ability to multitask and work with little direction in a collaborative fast paced environment. Strong problem-solving skills and ability to track software issues to successful resolution. Demonstrated skills to implement and integrate new technologies along with best practices. Excellent oral and written communication skills, and interpersonal communication skills with internal and external partners. Our ideal candidate would reside within the Bay Area, California, however, we would consider a remote working arrangement for the right candidate.