GPU Software Development Engineer

AMD; Calgary; On-site; USD 148,000 - 224,000; Full time

Job summary

Join a forward-thinking company as a GPU Software Development Engineer, where you'll contribute to high-performance computing and machine learning. This role involves developing and optimizing GPU-accelerated libraries that drive innovation in AI and scientific computing. You'll collaborate with talented mathematicians and engineers, delivering high-quality open-source code that enhances performance across diverse applications. If you're passionate about cutting-edge technology and eager to make a significant impact, this opportunity is perfect for you.

Benefits

Comprehensive benefits

Diversity and inclusion initiatives

Qualifications

• Strong background in C and C++ application and library development.
• Experience with GPU software development using HIP or CUDA.

Responsibilities

• Implement, optimize, and support GPU-accelerated collectives and algorithms.
• Benchmark, profile, and optimize code for throughput on single and multi-GPU systems.

Skills

C

C++

GPU software development

Machine learning

Numerical problem solving

Attention to detail

Performance improvement

Education

B.Sc. or B.Eng. in Computer Science

Advanced degrees (M.Sc., M.Eng., Ph.D.)

Tools

HIP

CUDA

Version control

Issue tracking

Job description

Join to apply for the GPU Software Development Engineer role at AMD.

This range is provided by AMD. Your actual pay will be based on your skills and experience — talk with your recruiter to learn more.

Base pay range

$148,720.00/yr - $223,080.00/yr

About AMD and the Role

AMD is seeking talented mathematicians, scientists, and engineers to develop math libraries that enable high-performance computing and machine learning workloads at Exascale. You will work on open-source, GPU-accelerated libraries such as rocPRIM, rocRAND, rocThrust, and hipCUB, which are part of the AMD ROCm Software platform.

The Libraries group in AMD AI GPU Software (AGS) builds GPU-accelerated, open-source math and communication libraries powering AI and high-performance computing applications on workstations, data centers, and supercomputers.

Candidate Profile

You should be comfortable working in dynamic, geographically distributed agile teams, with excellent communication skills, attention to detail, and a focus on performance improvement of software and hardware.

Key Responsibilities

• Implement, optimize, and support GPU-accelerated collectives and algorithms.
• Benchmark, profile, and optimize code for throughput on single and multi-GPU systems.
• Deliver high-quality code and documentation following open-source best practices.
• Collaborate with AMD experts, partners, and customers to enhance ROCm applications, libraries, and tools.

Preferred Experience

• Strong background in C and C++ application and library development.
• Experience with GPU software development using HIP or CUDA.
• Understanding of CPU and GPU architectures, low-level optimization, assembly programming, and vectorization.
• Experience solving numerical problems in machine learning and scientific computing.
• Knowledge of software development practices including testing, profiling, debugging, documentation, version control, and issue tracking.

Academic Credentials

• B.Sc. or B.Eng. in Computer Science, Software Engineering, Electrical Engineering, Applied Mathematics, or related fields.
• Advanced degrees (M.Sc., M.Eng., Ph.D.) are preferred.

Location

Calgary, Alberta

AMD offers comprehensive benefits. We are committed to diversity and inclusion, considering all qualified applicants without discrimination.