CUDA Kernel Engineer for ML Training & Inference

Lever, Inc.; Sunnyvale (CA); On-site; USD 150,000 - 450,000; Full time

Benefits offered by this job

Comprehensive medical, dental, and视觉 V

Bonus

401K Plan

Generous paid time off, sick leave and

Paid Parental Leave

Employee Assistance Program

Life insurance and disability

Job summary

MBZUAI is seeking a GPU Kernel Engineer to optimize machine learning software stacks for training and inference on state-of-the-art hardware. You will work on CUDA kernel development, profiling, and performance tuning, with opportunities to lead design reviews and contribute to documentation.

The role emphasizes hands-on system-level coding, parallel computing, and large-scale ML experience, with distributed training as a plus. Collaboration with the institute’s HPC teams is expected.

Qualifications

• Strong C++ and CUDA kernel development experience for deep-learning workloads.
• Proficiency in Python and integrating kernels with PyTorch or similar frameworks.
• Experience profiling and optimizing GPU kernels with performance-focused mindset.

Responsibilities

• Understand, profile, optimize and guide deep learning workloads on state-of-the-art hardware and software platforms.
• Design and implement performance benchmarks and testing methodologies.
• Build tools to automate workload analysis and optimization.

Skills

C++ programming

CUDA kernel development

Python with PyTorch integration

Tools

Triton

CUTLASS

PTX/SASS analysis

Job description

MBZUAI is seeking a GPU Kernel Engineer to optimize machine learning software stacks for training and inference on state-of-the-art hardware. You will work on CUDA kernel development, profiling, and performance tuning, with opportunities to lead design reviews and contribute to documentation.

The role emphasizes hands-on system-level coding, parallel computing, and large-scale ML experience, with distributed training as a plus. Collaboration with the institute’s HPC teams is expected.