[caption id="attachment_6906" align="alignnone" width="5000"] Image Credit: cdn.banyanhill.com[/caption]

Edge AI chips will help stimulate technology’s performance, enhancing data privacy. 

The capabilities of artificial intelligence (AI) across diverse industries are not concealed from anyone, delivering enhanced business productivity. However, as AI promises huge benefits to organizations, the future of this technology lies on the edge. AI at the edge will act similar as humans perform, where each learning from the surrounding environment will make locally optimal decisions. With the development of new chip technologies for edge AI, there will be no need to relay data to the central cloud infrastructures. This means using edge AI chips, a device will be able to assess data in real-time and send only relevant data for further analysis in the cloud. This process will certainly lessen the storage and bandwidth cost at all.

According to Deloitte predictions, over 750 million edge AI chips – chips or parts of chips that act or expedite machine learning tasks on-device, instead of in a remote data center – will be sold in 2020, accounting for a staggering US$2.6 billion in revenue. The vast majority of these edge AI chips will go into the consumer device market, especially into high-end smartphones.

In the context of biometric data, the chips promise better precision for biometric facial recognition algorithms. Recently, Nvidia, the leading provider of high-powered graphics processing units (GPU), announced to offer edge AI software, hardware for AI and biometrics to healthcare staff on the frontline. The company has new software and AI models that it expects will play a vital role in biometric applications in the healthcare industry during the COVID-19 pandemic and beyond.

Moreover, to enhance hospital public safety and patient monitoring, Nvidia is providing pre-trained models which consist of support for vision, speech and NLP using its EGX Edge AI hardware offerings. On the other side, ARM, a semiconductor company, is expanding its efforts with two new chip designs – the Arm Cortex-M55 and the Ethos-U55. The edge AI delivers numerous benefits including running AI processing on a device itself, rather than in a remote server, offers massive advantages to privacy and speed when it comes to handling these requests.

Smart Cities at the Edge AI

Today, the race for smart cities has become a business opportunity, and with the evolution of IoT, 5G and edge computing, among others, companies could derive exponential benefits for smart cities. The major enabler of the AI edge chipset is the emergence of Smart Cities initiatives worldwide. Dell, for instance, has been constantly marketing its capabilities in this market, and is joining the list of companies including AT&T, Verizon, Comcast, Nvidia, Intel, Qualcomm, Ericsson, Nokia, Cisco and others.

When it comes to ensuring privacy in smart cities, edge AI chips will be key that will analyze data while lessening privacy issues. But some critical data will still need to be relayed and stored in a central location. This is where 5G and other wireless technologies come into the rescue.