Senior ML Accelerator Engineer - GPU

Description

The AV organization builds advanced automated driving technology, including Level 4-capable fully self-driving systems, aimed at safer, more sustainable and more accessible mobility.

The AI Kernels & Compilers team turns perception, prediction and planning research into production software that runs efficiently and reliably on real vehicles at scale. The work covers model export, kernel development and performance engineering, with the goal of better situational awareness, faster reactions and more robust behavior on the road.

The AI Kernels team builds high-performance GPU kernels and custom libraries for on-vehicle ML inference in ADAS and autonomous driving. The team makes core AI workloads faster, more reliable and easier to maintain and deploy in cars under real-world constraints.

The team works closely with AI Solutions, AI Compilers, AI Architecture and AI Tooling so models deploy efficiently to the car and meet strict latency, throughput and reliability targets.

Responsibilities

• Write, benchmark and refine CUDA kernels and custom operators that speed up on-vehicle inference.
• Build tooling and infrastructure to profile, debug and validate CUDA kernels and accelerator backend code across the AV stack.
• Work with AI Solutions, Compilers and Architecture to turn model and system needs into kernel roadmaps, priorities and project plans.
• Work with compiler, performance tooling, runtime and deployment teams to ship reusable, reliable, high-performance libraries to production.
• Set standards, methods, processes and guidelines for GPU kernel development and performance engineering, applied through code review.
• Manage relationships with internal customers so kernels and libraries meet their real needs.

Requirements

• At least 2 years of relevant industry experience, or equivalent experience.
• A BS, MS or PhD in computer science or a related technical field.
• Strong CUDA GPU programming skills, with a solid grasp of parallel programming patterns and GPU architecture.
• Hands-on experience benchmarking, profiling, debugging and optimizing accelerator libraries and kernels with NSight tools or similar.
• Strong software architecture, library design and design pattern skills.
• Strong C++ skills and comfort working in large codebases.
• Solid background in system performance, high performance computing and/or architecture-aware optimization.
• Strong communication skills and the ability to work well in a team.
• Excellent analytical and problem-solving skills.
• Preferred: at least 2 years of relevant industry experience, or equivalent experience.
• Preferred: experience with tensor core programming, CUTLASS and/or CuTe.
• Preferred: experience with ML model architectures, particularly transformer-based ones.
• Preferred: experience with low latency or real-time systems.
• Preferred: experience with lower levels of an accelerator software stack, such as drivers, runtimes and compilers.

Conditions

Pay: the base salary range is $170,100 to $258,300. The actual offer falls within this range and depends on job-related factors. The compensation figure is a good faith estimate based on what a successful applicant might be paid under applicable state law, and it may not reflect roles located outside New York, Colorado, California or Washington.

Bonus: an incentive pay program pays out based on company performance, job level and individual performance.

Hybrid: you are expected to be on site at a specific location at least 3 days a week, or as your manager sets.

Travel: up to less than 25% of the time.

Relocation benefits are available to candidates who qualify under company policy.

The company is an equal opportunity employer and bases decisions on non-discriminatory criteria.

Benefits

• Health and wellbeing programs including medical, dental, vision, Health Savings Account, Flexible Spending Accounts, retirement savings plan, sickness and accident benefits, life insurance, paid vacation and holidays, tuition assistance, employee assistance program and vehicle discounts.
• Support for your wellbeing at work and at home from your first day, with details on the company's total rewards available online.