AUSTIN — Ambiq Micro, Inc., a technology leader in ultra-low power semiconductor solutions for edge AI, today announced it has won the 2026 Embedded World Award in the Best Tools category for heliaAOT™, its ahead-of-time AI compiler purpose-built for Ambiq’s ultra-low power SoCs. heliaAOT was also nominated in the Best AI category. This marks the third consecutive year Ambiq has received a prestigious Embedded World Award, underscoring the company’s sustained industry leadership.
As AI workloads rapidly migrate from cloud to edge devices, developers face mounting constraints in power, memory, and compute. Introduced in July 2025, heliaAOT addresses this challenge by converting TensorFlow Lite models directly into highly optimized embedded C code — eliminating runtime overhead and enabling breakthrough efficiency on battery-powered devices.
Key advantages include:
- 10x reduction in memory footprint versus traditional runtime-based deployments
- Granular, per-layer weight distribution, control across Apollo’s memory hierarchy
- Deterministic, automated memory planning that eliminates over-allocation
- Streamlined deployment with direct integration of generated C code into embedded applications
- Seamless support through Zephyr RTOS and Ambiq neuralSPOT® AI SDK
“We are deeply honored by the recognition from Embedded World 2026,” said Dr. Adam Page, Head of Artificial Intelligence at Ambiq. “heliaAOT is a strategic investment in the future of edge AI—addressing extreme resource constraints while meeting the rising demand for agentic, always-on intelligence. Together with our SPOT platform, heliaAOT delivers a powerful hardware-software advantage with orders-of-magnitude energy efficiency and seamless deployment.”
Additional information and benchmarks for heliaAOT are available here.
