Senior Software Engineer (Architecture)

Targeted Talent; Quebec; Hybrid; CAD 120,000 - 160,000; Full time

Job summary

A technology company specializing in AI solutions is seeking a Software Architect to lead software development for a high-performance AI inference chip. Responsibilities include advancing compiler technology, optimizing software for parallel architectures, and collaborating with hardware engineers. The ideal candidate has over 10 years of relevant experience, strong skills in C/C++ and Python, and a background in Deep Learning. The role offers competitive benefits, including stock options and 20 vacation days.

Benefits

20 vacation days

Strong health and extended health benefits

Unlimited sick days

Stock options

Qualifications

• 10+ years of experience developing software for highly parallel architectures.
• Experience optimizing algorithms for hardware acceleration.
• Strong problem-solving skills to resolve complex issues.

Responsibilities

• Advance compiler and runtime technology for AI workloads.
• Research and design new software and hardware solutions.
• Evaluate trade-offs of different parallelization strategies.
• Enhance AI software tools to support DNNs.
• Collaborate with engineers to develop deep learning software.

Skills

Software development for highly parallel architectures

Optimizing algorithms for hardware acceleration

Problem-solving skills

Deep Learning fundamentals

Development in C/C++

Development in Python

Soft skills in a group environment

Education

Master's or PhD in Computer Science, Engineering, or related field

Job description

About the Company:
Our client is a company building the world's highest- performance pure digital AI inference chip.
They are seeking a Software Architect to lead their software efforts and advance the software stack that includes ML frameworks, compilers, libraries, and runtime. As a Software Architect, you will be responsible for designing and developing software that interacts with their chip, collaborating with hardware and software engineers, and optimizing software for highly parallel architectures to maximize performance and efficiency.

Responsibilities:

• Advance the state of the art in compiler and runtime technology for delivering high-performance acceleration of AI workloads across a variety of neural network architectures.
• Research and design new software and hardware AI solutions, involving simulators, optimizing compilers, code generators, and runtime execution frameworks for deep learning accelerators.
• Evaluate various trade-offs of different parallelization strategies such as performance, power, energy, and memory consumption.
• Enhance AI software tools to support the latest and greatest DNNs emerging from the research community and industry.
• Keep up with the fast-paced development happening in the industry and academia to continuously enhance our products.
• Work closely with other software and hardware engineers to develop the next generation of deep learning software.
• Collaborate with architects and hardware engineers to co-design future accelerators.

Preferred Skills & Experience:

• 10+ years of experience developing software for a variety of highly parallel architectures.
• Previous roles include experience with optimizing algorithms for hardware acceleration machine learning accelerators, spatial architectures, or GPUs.
• Strong problem-solving skills and the ability to resolve complex issues with a high level of ambiguity.
• Understanding of Deep Learning fundamentals.
• Strong development skills in C/C++, Python.
• Superb soft skills: ability to work efficiently and effectively in a group environment, and influence a cross-functional team without having direct managerial authority.
• Computer Science, Engineering, or related degree; preferably MS or PhD.
• Driven and self-directed.

Perks:

• 20 vacation days
• Strong health and extended health benefits
• Unlimited sick days
• Stock options

Please apply even if you don't check all the boxes. We are an inclusive and diverse company and welcome applicants from all backgrounds.