|An Android mascot posing in front of the Google headquarters. / Photo by: Asif Islam via Shutterstock|
Google made an announcement recently about its new edge products to better handle data on Day 2 of the online search giant's Cloud Next conference. This is in line with the growing influence and application of the internet of things in commercial and industrial settings that are now inundating businesses with increasing amounts of data.
According to the research and analysis by IDC, data bits double in size every two years and by 2020, there will be 44 trillion gigabytes created and copied in the digital universe. For Google, integrating machine learning to the edge -- a network where connected devices meet -- may work to easily and quickly manage the voluminous data.
Google created three new products: Edge TPU chip, Cloud IoT Edge software, and a developmental kit for Edge TPU designed to enhance big data management and decrease adverse effects of real-time analytics.
1. Edge TPU chip: It is a purpose-built application-specific integrated circuit or ASIC chip that will run AI at the edge. It has a small physical and power footprint while exhibiting high-performance. As part of its design, the chip can deploy algorithms and AI-associated software functions.
2. Cloud IoT edge: This cloud service enables the data processing and machine learning of Google Cloud to reach billions of different devices connected at the edge, such as robotic arms, oil rigs, and wind turbines. The service can run on Linux-based systems and devices running the Android Things platform.
3. Edge TPU development kit: A development kit for Edge TPU has been made to jump-start the development and testing of the ASIC chip. The kit has a system-on-module design to combine the Edge TPU, an NXP CPU, Wi-Fi, and the microchip secure parts in a compact form.
"Edge TPU will bring a brain to your embedded devices at extremely low cost and power efficiency without compromising performance. I believe this is a game changer," said Injong Rhee, vice president of IoT, Google Cloud.