Job Summary
The role involves spearheading the strategy for inference performance, aiming to enhance token efficiency and maximize GPU utilization in large-scale models. This position is crucial for developing benchmarking frameworks, tailoring inference engines, and applying advanced techniques in parallelism and kernel-level optimization.
Responsibilities
- Lead the optimization of inference performance to improve token efficiency and GPU utilization.
- Develop and implement benchmarking frameworks for performance assessment.
- Customize inference engines to meet specific performance requirements.
- Execute advanced parallelism techniques and kernel-level optimizations.
Qualifications
- A minimum of 6 years of experience in systems or performance engineering.
- 2-3 years of specialized experience in large language model (LLM) or GPU performance.
- A bachelor’s degree in Computer Science or a relevant field.
- Strong expertise in Transformer inference and CUDA kernel optimization.
Zenlayer, Inc provides equal employment opportunities to all employees and applicants for employment and prohibits discrimination and harassment of any type without regard to race, color, religion, age, sex, national origin, disability status, genetics, protected veteran status, sexual orientation, gender identity or expression, or any other characteristic protected by federal, state or local laws.
