ML Inference Engineer - Custom Accelerator Data Plane

Amazon; Cupertino (CA); On-site; USD 165,000 - 224,000; Full time

Job summary

Amazon Cupertino is seeking a Software Development Engineer for the MLIL DataPlane team to own design and implement inference data plane for large models on custom hardware. You will work on model execution, memory management, data movement, and serving integration.

You will develop and optimize compute kernels for a custom ML accelerator, validate architectures end-to-end, and build profiling infrastructure while driving performance across the stack from simulation to production on GPUs/ASICs.

Qualifications

• Bachelor's degree or equivalent.
• 4+ years of full software development life cycle experience.
• Knowledge of computer architecture, operating systems, and parallel computing.
• Knowledge of Linux fundamentals.
• Strong proficiency in C/C++.
• Experience developing compute kernels for GPUs, DSPs, or custom accelerators.
• Proven track record of owning and delivering complex software features end-to-end.

Responsibilities

• Develop and optimize compute kernels for a custom ML accelerator architecture, targeting production-level performance for large language model inference.
• Implement and validate LLM architectures end-to-end - from PyTorch model definition through distributed execution on custom hardware.
• Integrate custom accelerator backends into open-source ML serving frameworks (vLLM, PyTorch), including scheduler extensions, memory management, and model parallelism.
• Build and maintain test infrastructure for model correctness validation across CPU, GPU, simulator, and hardware targets.
• Profile and optimize inference workloads - identify bottlenecks, instrument critical paths, and drive latency and throughput improvements from simulation through hardware bring-up.
• Own features end-to-end: from design through implementation, testing, and integration into the broader software stack.
• Contribute to CI/CD pipelines that gate model and kernel changes on correctness and performance regressions.

Skills

C/C++

Linux fundamentals

Parallel computing

Computer architecture

Software development lifecycle

Education

Bachelor's degree or equivalent

Job description

Amazon Cupertino is seeking a Software Development Engineer for the MLIL DataPlane team to own design and implement inference data plane for large models on custom hardware. You will work on model execution, memory management, data movement, and serving integration.

You will develop and optimize compute kernels for a custom ML accelerator, validate architectures end-to-end, and build profiling infrastructure while driving performance across the stack from simulation to production on GPUs/ASICs.