Principal Edge AI Firmware Engineer
Ambiq Micro, Inc.2 months ago
Responsibilities
- Develop and support HeliaRT and HeliaAOT embedded AI runtimes with emphasis on portability, correctness, performance, and usability.
- Implement and optimize ML operator kernels and embedded libraries for DSP, vector, and NPU acceleration, including HeliaDSP and HeliaCore.
- Build on-device profiling and performance-analysis tools using PMU counter data.
- Optimize latency, memory footprint, bandwidth, and energy consumption through compute and memory-hierarchy analysis.
- Develop benchmark harnesses, microbenchmarks, and regression tests for numerical correctness and performance stability.
- Enable customer integration across bare-metal, FreeRTOS, and Zephyr environments and maintain documentation, guides, examples, and benchmarks.
- Improve runtime memory planning and efficiency while managing upstream and fork health.
Requirements
- Bachelor’s degree in Electrical Engineering, Computer Engineering, Computer Science, or a related field, plus 12+ years of relevant experience, or equivalent practical experience.
- Strong experience designing, developing, and testing embedded software in C/C++.
- Strong debugging practices focused on correctness, reproducibility, and preventing performance regressions.
- Understanding of compute, memory, bandwidth, cache effects, deterministic latency, and energy efficiency in constrained systems.
- Ability to interpret datasheets and reference manuals; schematic experience is a plus.
- Proficiency with Git or equivalent version control.
- Experience with Arm toolchains, GDB, J-Link or SEGGER-class probes, and CMake or Make build systems.
- Python experience for profiling, analysis, automation, or developer utilities is preferred.
- Rust experience is preferred.
- Familiarity with TensorFlow Lite for Microcontrollers or similar embedded inference runtimes is preferred.
- Experience optimizing embedded DSP, vector-extension, or NPU targets is preferred.
- Familiarity with CMSIS-NN, CMSIS-DSP, Vela, quantized inference, CMSIS-Pack, or CPM.cmake is preferred.
- Lab measurement experience involving power or debugging with basic bench tools is preferred.
- Must be currently authorized to work in the United States for any employer.
About Ambiq Micro, Inc.
Ambiq Micro is a fabless semiconductor company that designs ultra-low-power microcontrollers and AI-capable SoCs for wearables, IoT, and other battery-powered devices. Its Apollo MCU family and subthreshold power-optimized technology target always-on sensing and edge inference, with reference software and tools for OEMs and ODMs. Founded in 2010 and headquartered in Austin, Texas, it is privately held and serves consumer, industrial, and healthcare device makers.