AnalogAI Selects Microchip's SST memBrain SAGE IP to Build First Real-World Edge AI Solution

AnalogAI Selects Microchip's SST memBrain™ SAGE IP to Build First Real-World Edge AI Processor

— Proprietary On-Chip Training Technology Enables Dynamic Adaptation to Real-World Environments

AnalogAI has selected the memBrain™ Synaptic Analog Generation Engine (SAGE) neuromorphic hardware intellectual property (IP) from Silicon Storage Technology® (SST®), a subsidiary of Microchip Technology Inc., to develop edge AI processors for real-world scenarios. AnalogAI edge AI processors feature hardware-adapted proprietary algorithms that can simultaneously perform AI model training and inference tasks in real-world environments.

AnalogAI is using SST's memBrain SAGE IP to achieve high-performance analog compute-in-memory (aCIM) at power levels of 1 watt and below, targeting ultra-low-power edge applications. Leveraging the silicon-proven memBrain SAGE IP based on SST SuperFlash® technology, AnalogAI is developing inference applications for real-world scenarios including humanoid robots, drones, and automotive devices.

"AnalogAI is committed to developing innovative, co-optimized edge AI inference solutions that enable edge devices to adapt in real time to unexpected environmental changes," said Mark Reiten, Senior Vice President of Microchip's Intelligent Computing Business Unit. "As the core inference engine in AnalogAI's first products, SST memBrain SAGE IP delivers the computational performance and energy efficiency their applications require. We warmly welcome AnalogAI as the latest licensing partner of SST memBrain IP, joining our growing ecosystem."

Key features of SST memBrain SAGE IP:

· memBrain Tensor In-Memory Logic Element (TILE)

o The memBrain solution uses optimized ESF3 memory cells, achieving up to 8 bits/cell storage capability at nanoamp-level power consumption

o The memBrain solution features custom arrays, decoders, and driver circuits

· Optimized digital-to-analog converters (DAC) and analog-to-digital converters (ADC)

· Summer and high-voltage bias circuits

· Nanoamp-level memory cell control logic

· memBrain proprietary test circuits

· SST memBrain technical documentation and simulation models

· Full suite of IP integration services and technical support

"After evaluating available options across the industry, we ultimately chose SST's memBrain SAGE IP," said Jaejun Lee, CEO of AnalogAI. "This silicon-proven IP helps us shorten development cycles while meeting the dual requirements of ultra-low power and high performance in our target market segment. Analog compute-in-memory is a rapidly growing emerging field, and AnalogAI is developing new capabilities for edge AI devices that require real-world environmental adaptability."

SST's memBrain SAGE IP has been developed and deployed on production-ready SuperFlash® memory using 40 nm and 28 nm foundry processes; the current technology roadmap also includes development of 22 nm memBrain SAGE IP. SuperFlash memory enables highly reliable, high-performance, low-power non-volatile storage on-chip, making it broadly suitable for applications requiring fast access, high endurance, and long-term data retention without the need for external memory devices.

Customers interested in learning about SST memBrain IP solutions and SuperFlash technology are invited to visit the SST website or contact their regional SST sales manager for details. For information about AnalogAI products, please visit the AnalogAI website or email contact@analog-ai.com.

About Microchip Technology Inc.

Microchip Technology Inc. is a leading provider of a rich portfolio of semiconductor products, dedicated to making innovative design easier through complete system-level solutions that help customers address key challenges in bringing emerging technologies to mature markets. The company offers easy-to-use development tools and a broad product portfolio to support customers throughout the design process from concept to final implementation. Headquartered in Chandler, Arizona, USA, Microchip continues to deliver solutions for industrial, automotive, consumer, aerospace and defense, communications, and computing markets through exceptional technical support. For more information, please visit the company website at www.microchip.com.

SST (Silicon Storage Technology), a subsidiary of Microchip, is a leading provider of embedded flash memory technology. SST develops, designs, licenses, and markets a portfolio of proprietary and patented SuperFlash memory technology solutions for the consumer, industrial, automotive, and Internet of Things (IoT) markets. Founded in 1989, SST went public in 1995 and was acquired by Microchip in April 2010. SST is now a wholly owned subsidiary of Microchip, headquartered in San Jose, California. For more information, please visit the SST website at www.sst.com.

AnalogAI is dedicated to developing edge AI processors capable of on-device training, enabling robots, drones, and vehicles to adapt in real time to changing physical environments while operating at ultra-low power. Its analog compute-in-memory architecture enables both on-device training and inference in a single low-power design, breaking through the limitation of conventional edge AI hardware that supports inference only. AnalogAI focuses on edge applications for the physical world, enabling systems to continuously adapt to environmental changes after deployment, with applications spanning robotics and autonomous machines. The company is headquartered in Seoul, South Korea. For more information, please visit www.analog-ai.com.