Our SMU-E™️ product is a standard CMOS logic-compatible, cost-effective, non-volatile processor to empower your real-time edge intelligence of running TinyML with significantly higher performance per given battery power. The development board is available along with a pre-configured software development kit for simple deployment.
Conventional MCU is very expensive in the advanced process node due to the extra masks and steps beyond standard CMOS. We provide a unique standard CMOS logic-compatible non-volatile processor solution integrating Logic-EFLASH and Compute-in-Flash (CiF) based energy-efficient neural engine for cost and performance effective TinyML acceleration.
We support optimized referencing of TinyML neural network models on the smart edge devices with SMU neural engine.
We provide cross-platform SW development toolchain and SMU neural engine converters for pre-trained ML models.
Our smart microprocessor features standard logic compatible Embedded FLASH (Logic-EFLASH) and Compute-in-Flash (CiF) based energy efficient neural engine suitable for battery powered TinyML applications.