Rising edge AI requirements demand solutions

IHS expects over 40 billion devices will be operating at the network Edge by 2025, while market intelligence firm Tractica predicts over 2.5 Billion Edge devices will ship annually by the same year.

As a new generation of Edge applications emerges, designers are increasingly pressed to develop solutions that combine low power and small form factor without compromising performance. Driving demand for these new AI solutions are a growing range of Edge applications such as presence detection for home control in applications like smart doorbells and security cameras, object counting for retail applications like inventory, and object and presence detection in industrial applications.

On one hand, the market demands that designers develop solutions that operate at performance rates higher than ever before. On the other hand, latency, bandwidth, privacy, power and cost issues constrain them from relying on computational resources in the cloud to perform analytics.

At the same time, performance, power and cost limitations vary enormously from application to application. As data requirements for always-on Edge applications continue to drive demand for cloudbased services, designers must address traditional concerns about power, footprint and cost. How can developers address rising system constraints on power budgets that run in the mW range and form factors that can range from 5mm² to 100mm². The wide range of performance requirements alone is difficult to track.

Lattice’s FPGAs are uniquely positioned to address the rapidly changing market for Edge devices. One way designers can quickly bring more computational resources to the Edge without relying on the cloud is to use the parallel processing capabilities inherent in FPGAs to accelerate neural network performance.

Moreover, by using lower density FPGAs optimized for low power operation and available in compact packages, designers can meet the stringent power and footprint limitations associated with new consumer and industrial applications.

As an example, Lattice’s iCE40 UltraPlus and ECP5 product families support development of Edge solutions that consume anywhere from 1mW to 1W on compact hardware platforms ranging from 5.5mm² to 100mm². 

By combining ultra-low power, high performance and accuracy with comprehensive legacy interface support, these FPGAs give Edge device developers the flexibility they need to address changing design requirements.

To read more, download the whitepaper.