AI Accelerator Software Principal Engineer - NPU Full-Stack Integration

Ampere Computing

Portland, OR · RemoteFull-time$182,000 – $273,000 / yearly

Apply now

Job details

Pay

$182,000 – $273,000 / yearly

Job type

Full-time

Work setting

Remote / Hybrid

Location

Portland, OR

Posted

Posted 2 months ago

Full job description

Lead the design and delivery of high-performance, low-latency deep learning inference solutions on the Arm Ethos-U85 at Ampere. Advance the AI software stack at the intersection of software engineering, performance engineering, and hardware-aware optimization. Hybrid role in Portland, OR.

What you'll do:

Execute end-to-end deep learning performance acceleration across the full stack: framework integration, compiler support, runtime libraries, compute kernels
Improve performance and accuracy for production-ready inference on edge devices
Co-optimize AI execution with hardware and platform teams for throughput, latency, and scalability
Develop AI co-processor and accelerator software with reusable libraries and optimized execution paths
Work cross-functionally with compiler, runtime, kernel, platform, and product teams
Profile, benchmark, and tune systems for realistic workloads

Pay range: $182,000 - $273,000.

Requirements

BS in CS, Engineering, or related field with 8+ years experience (MS with 6+, PhD with 3+)
Understanding of AOT compilation paths (PyTorch) and deployment tools (executorch)
Expert proficiency in Python and C/C++
Strong performance profiling and tuning background
Solid understanding of AI/ML concepts including neural networks
Experience with Transformers and Diffusion models preferred

SemiconductorAI

Ready to apply?

Apply on the company site

You’ll be taken to an external site to complete your application.