Software Engineer Specializing in GPU Performance and Deep Learning

About the Role

Drive innovation in deep learning and GPU performance as a Software Development Engineer. Collaborate within a dynamic team to optimize frameworks and enhance AI experiences across platforms.

In this pivotal role, you’ll focus on optimizing deep learning frameworks and developing GPU kernel solutions to accelerate AI model inference and training. Your analytical expertise in C++ and experience with open-source technologies will foster contributions that enhance AMD’s AI ecosystem. Collaborate with internal teams to drive efficiency and performance improvements across multi-GPU infrastructures.

Key Responsibilities

• Optimize deep learning frameworks like PyTorch for GPUs
• Implement and refine GPU kernels for efficiency
• Collaborate on compiler-level optimizations with internal teams
• Support the growth of SGLang across AMD’s GPU offerings
• Tune performance for distributed multi-GPU environments

Requirements

• Strong programming skills in C++ and/or Python
• Hands‑on experience with GPU programming models
• Knowledge of optimization in heterogeneous environments
• Experience with AI frameworks integration
• Degree in Computer Science or equivalent

Additional Information

Contribute to cutting‑edge AI solutions by optimizing frameworks and enhancing GPU capabilities, fostering innovation for the future.

#J-18808-Ljbffr