Staff Engineer

Company:

Qualcomm India Private Limited

Job Area:

Engineering Group, Engineering Group > Systems Engineering

General Summary:

What is the product you will be working on?

You will be working on Qualcomm® Cloud AI 100, the industry leader in performance/watt and performance density. It is purpose-built for high performance, low-power AI processing in the cloud. It is built from the ground up to help accelerate AI experiences, providing a cutting-edge solution that addresses the most important aspects of cloud AI inferencing—including low power consumption, scale, process node leadership, and signal processing expertise.

What is the team you will be joining?

You will be joining as a deep learning architect to the team that has following mission:

• HW architecture of the SoC components that are related to performance in the current and next-generation accelerators.
• Roofline performance analysis of customer networks on current and next-generation accelerators.
• Keeping abreast of developments in emerging research in the industry and academia to shape the next-generation accelerators. 

Minimum Qualifications:

Minimum Qualifications
• Ph.D in Computer Science or related field (or) Master's degree in Computer Science or related field with related work experience.
• Track record of few publications in a relevant field.
• Experience in frameworks and formats (PyTorch, Tensorflow, ONNX)
• Ability to write novel neural network architectures in these frameworks.
• Ability to comprehend complex implementations available in HuggingFace-like repos.
• Ability to modify existing implementations available in HuggingFace-like repos.
• Deep understanding of proven NN architectures and foundational theories in NLP, Vision, Graphs
• Good understanding AI Software tools on well-established accelerator in the industry.
• Last but not least, passion to continuously learn and discuss new ideas.

Preferred Qualifications
• Good understanding of the AI Software tool chains.
• Ability to compile and run the models on Accelerators or GPU
• Ability to collect statistics and analyze them to identify bottlenecks.
• Ability to Quantize and experience with Qualcomm AIMET tools is a plus.

• You will analyze customer networks and understand the various architectural building blocks. This will involve reviewing large Pytorch or Tensorflow repositories or an ONNX representation and requires deep understanding of the proven neural network architectures in NLP, Vision and Graphs.
• You will think about how they can be mapped optimally into our current and next-gen HW architectures. This will require in-depth understanding of the HW capabilities of current and next-gen architectures.
• Some of the customer networks are architected to be optimal on accelerators with different structure than our accelerators. You will propose ideas to rearchitect them with our accelerators as target and implement them. This involve rewriting sections of model in Pytorch/Tensorflow , and evaluating the task-appropriate accuracy metrics on task-appropriate datasets.
• Optionally, you will also validate the performance of the rearchitected neural network on our accelerators. This will require in-depth understanding of the QCOM AI SW tool chains and involves collaborating with the onboarding team, compiler team , and performance evaluation teams.
• You will track bleeding-edge research on NLP, Vision, Graphs etc., as they appear on arxiv,  judiciously deep dive into the details of some of them and present the new techniques to others in the team. Also,  you will look for opportunities apply the techniques to improve current practices and shape the HW architecture of next-generation HW architectures.

Applicants: If you need an accommodation, during the application/hiring process, you may request an accommodation by sending email to accommodationsupport 

Although this role has some expected minor physical activity, this should not deter otherwise qualified applicants from applying. If you are an individual with a physical or mental disability and need an accommodation during the application/hiring process, please call Qualcomm’s toll-free number found here for assistance. Qualcomm will provide reasonable accommodations, upon request, to support individuals with disabilities as part of our ongoing efforts to create an accessible workplace.

Qualcomm is an equal opportunity employer and supports workforce diversity.

Qualcomm expects its employees to abide by all applicable policies and procedures, including but not limited to security and other requirements regarding protection of Company confidential information and other confidential and/or proprietary information, to the extent those requirements are permissible under applicable law.

To all Staffing and Recruiting Agencies: Our Careers Site is only for individuals seeking a job at Qualcomm. Staffing and recruiting agencies and individuals being represented by an agency are not authorized to use this site or to submit profiles, applications or resumes, and any such submissions will be considered unsolicited. Qualcomm does not accept unsolicited resumes or applications from agencies. Please do not forward resumes to our jobs alias, Qualcomm employees or any other company location. Qualcomm is not responsible for any fees related to unsolicited resumes/applications.

If you would like more information about this role, please contact Qualcomm Careers.