Senior Software Engineer (GPU)

Quick Facts

• Domain: Semiconductors / advanced AI systems
• Focus: Low-level, ultra low-latency GPU and AI accelerator software

Description

You will work on low-level software for ultra low-latency GPU processors and AI accelerators. You’ll collaborate with hardware teams, system architects, and compiler engineers to improve performance and behavior of future compute platforms. Responsibilities include designing, implementing, optimizing, profiling, and benchmarking deep-learning algorithms and kernels.

Responsibilities

• Design, implement, and optimize low-level software implementing deep learning algorithms for ultra low-latency processors
• Work closely with hardware engineers, system architects, and compiler teams to integrate software with new hardware features
• Profile, benchmark, optimize, and tune kernels for optimal resource usage and minimal latency
• Write clear, maintainable C/C++ and Python code, and develop internal libraries and infrastructure

Requirements

• Very strong C/C++ knowledge for performance-critical systems
• Documented experience creating high-performance kernels for GPUs or dedicated accelerators
• Advanced optimization techniques: vectorization, memory hierarchy management, parallel processing, and effective cache usage
• Ability to read and understand complex codebases and algorithms
• Strong problem-solving and debugging skills
• Ability to communicate effectively and work in a team
• Technical bachelor’s degree
• English B2+ for international collaboration

Benefits

• Long-term employment contract directly with the client; open to candidate proposals with a flexible budget