Ambiq Micro announced heliaCORE, the production-grade AI kernel foundation for its HELIA Edge AI ecosystem, on June 23, 2026. Built specifically for Apollo SoCs, the software helps developers validate and deploy optimized edge AI products at scale. It arrives as the industry grapples with moving AI from experiments to real-world applications on resource-constrained devices, particularly in consumer electronics and wearables. heliaCORE is available now to Ambiq customers and partners.
What heliaCORE delivers for Apollo SoCs
heliaCORE provides optimized AI kernels, broad operator acceleration, and production-scale validation for Ambiq's low-power platforms. It accelerates not only core neural network kernels but also critical inference operations - including activations, padding, reductions, reshapes, and data movement - to improve real-world performance and efficiency. The software is built on the Arm CMSIS-NN ecosystem, preserving compatibility with standard embedded AI development workflows.
Validation across 40+ models and 963 operator instances
The solution was validated across more than 40 AI models, 53 operator types, 247 unique operators, and 963 operator instances. This validation footprint spans approximately 12 times more operator instances than MLPerf Tiny, providing broader coverage of real-world AI workloads. heliaCORE includes more than 200 accelerated operators optimized for Apollo devices, with MVE and DSP acceleration across current- and previous-generation Apollo SoCs.
"The future of AI will be defined by intelligent devices operating at the edge, but moving AI from experimentation to real-world deployment remains one of the industry's biggest challenges," said Dr. Adam Page, Head of AI at Ambiq. "heliaCORE is a foundational building block of the HELIA ecosystem, providing the software infrastructure developers need to deploy edge AI faster and with greater confidence."
Integration into existing workflows
heliaCORE is available through multiple integration flows:
- CMake
- CMSIS-Pack
- Zephyr
- neuralSPOT-X
This lets development teams incorporate it into their toolchains with minimal effort. For product development teams working on resource-constrained designs, the software reduces integration effort and accelerates time to market, a key focus in AI for Product Development. Developers can access documentation, software resources, and get-started materials through Ambiq's developer portal and the heliaCORE GitHub page.
Why this matters for product development
Product developers building AI-enabled devices on ultra-low power platforms now have a validated, production-ready software layer that cuts integration time and de-risks deployment. The 12x increase in operator coverage over MLPerf Tiny reduces the chance of unexpected failures when moving from prototype to production. Support for current and previous Apollo SoC generations means teams can extend existing hardware investments while accelerating their AI feature roadmap.
Your membership also unlocks:
