Maxim Integrated Products has released the MAXREFDES178# camera cube reference design, demonstrating how AI applications previously restricted to machines with large power and cost budgets may be embedded in space-constrained, battery-powered edge devices. The design facilitates ultra-low-power IoT devices to implement hearing and vision and showcases the MAX78000 low-power microcontroller with neural network accelerator for audio and video inferences. The system also comprises the MAX32666 ultra-low-power Bluetooth microcontroller and two MAX9867 audio CODECs. The complete system is provided in an ultra-compact form factor to present how AI applications such as facial identification and keyword recognition may be embedded in low-power, cost-sensitive applications such as IoT devices and wearables.
AI applications demand intensive computations, usually performed in the cloud or inexpensive, power-hungry processors that only fit in applications with big power budgets, such as self-driving cars. But the camera cube shows how AI can live on a low-power budget, allowing applications that are time and safety-critical to work on even the smallest of batteries. The MAX78000’s AI accelerator cuts the power of AI inferences up to 1,000x for vision and hearing applications when compared to other embedded solutions. The AI inferences running on the design also display dramatic latency improvements, running more than 100x faster than on an embedded microcontroller.
The compact size of the camera cube at 1.6" x 1.7" x 1.5" (41mm x 44mm x 39mm) shows that AI can be implemented in wearables and other space-constrained IoT applications. The MAX78000 solution itself is up to 50% smaller than the next-smallest GPU-based processor. It does not need other components such as memories or complex power supplies to implement cost-effective AI inferences.
“The next big opportunity in AI is providing machine learning insights at the edge,” said Alan Descoins, CTO at Tryolabs. “The MAXREFDES178# shows how Maxim Integrated’s AI solution is a breakthrough in power, latency and size that can unlock the possibilities for AI in battery-powered designs.”
“Machine learning promises a lot: that machines can make sense of what they see and hear like humans, as well as make more autonomous decisions. Until the MAX78000, the embedded world was left behind because you couldn’t implement AI at the edge in a power, cost and size constrained manner,” said Kris Ardis, executive director of the Micros, Security and Software Business Unit at Maxim Integrated. “Now the MAXREFDES178# demonstrates how meaningful and powerful AI inferences can be run at the edge, on even the smallest and most energy-conscious devices.”