Engineer - Machine Learning Software - Qualcomm

Polluxa, Inc.; California (MO); On-site; USD 90,000 - 130,000; Full time

Job summary

Polluxa, Inc. in the United States is seeking a software engineer to build high performance C++17 code for Qualcomm Hexagon processors in an embedded AI stack.

You will design and optimize a modern C++17 library, work with cross-compiler toolchains for Android, Linux, and QNX, and develop model analysis tools for internal and external customers. This replacement position offers collaboration across a globally distributed team and emphasis on software quality.

Qualifications

• Bachelor's degree in Computer Science, Engineering, Information Systems, or related field.
• 3+ years of relevant work experience.
• Experience with modern C++17 language features.
• Background in mathematical algorithms using fast math libraries and vector instructions.
• Practical experience with developing middleware or firmware software.
• Experience with multitasking and multithreading driver development.
• Experience with Hexagon DSP SDK or cross compiler toolchains for embedded systems.
• Familiarity with TensorFlow, PyTorch or ONNX.
• Experience with tools such as git, Linux, JIRA and Docker.
• Knowledge of design patterns.

Responsibilities

• Development of modern C++17 software library for Qualcomm Hexagon Processors.
• Design and performance tune modern C++17 code for an embedded system.
• Development of model analyzing tools for the internal and external customers.
• Use cross compiler toolchains for embedded systems such as Android, embedded Linux and QNX.
• Address issues found in existing and past Qualcomm AI products.
• Implement and optimize modern C++17 machine learning operations on Hexagon Processors.
• Debug customer machine learning use cases executing on Qualcomm AI Stack.
• Communication across globally diverse team.
• Participate in software quality process improvements.

Education

Bachelor's degree in Computer Science, Engineering, Information Systems, or related field

Tools

Hexagon DSP SDK

Cross compiler toolchains

Git

Linux

JIRA

Docker

Job description

Company:

Job Area:

General Summary:

AI's ability to solve complex problems across multiple domains is transformative. In the AI Software team, we build the Qualcomm AI Engine direct to enables OEMs and developers to run their deep neural network (DNN) models on Qualcomm Hexagon Processors. Our team works with OEMs and developers to develop and optimize DNN models for the Qualcomm AI Stack. We are building optimized on-device AI stack with cutting edge hardware to run deep neural networks with that phone you keep in your pocket, that car you drive, or that vacuum cleaner you unleash to clean your house. Come join us if you want to work on bleeding edge AI technology. In this position you will build high performance software for AI engines to extend our AI solutions into industry leading customer use cases. Replacement Position

Duties and Responsibilities

Development of modern C++17 software library for Qualcomm Hexagon Processors

Design and performance tune modern C++17 code for an embedded system

Development of model analyzing tools for the internal and external customers

Use cross compiler toolchains for embedded systems such as Android, embedded Linux and QNX

Address issues found in existing and past Qualcomm AI products

Implement and optimize modern C++17 machine learning operations on Hexagon Processors

Debug customer machine learning use cases executing on Qualcomm AI Stack

Communication across globally diverse team

Participate in software quality process improvements

Preferred Qualifications:

Three or more years of relevant work experience

Experience with modern C++17 language features

Background in mathematical algorithms using fast math libraries and vector instructions sets

Practical experience with developing middleware or firmware software

Experience with multitasking and multithreading driver development

Experience with Hexagon DSP SDK or cross compiler toolchains for embedded systems

Familiarity with TensorFlow, PyTorch or ONNX

Experience with tools such as git, Linux, JIRA and Docker

Knowledge of design patterns

Minimum Qualifications:

• Bachelor's degree in Computer Science, Engineering, Information Systems, or related field.