Alibaba Alibaba Cloud Apsara Conference Smart City

Alibaba Cloud to Upgrade Smart City With Edge Computing and AI Chip Hanguang 800

Rebbeca Ren

posted on September 29, 2019 7:00 am

September 25, Chinese tech behemoth Alibaba Group's Apsara Conference was held in Hangzhou. Yu Pengwu, Vice President of the group, unveiled the technical layout and future utilization of Alibaba Cloud, the cloud arm of the tech powerhouse, in the edge computing field.

Alibaba Cloud inaugurates its edge computing business from 2016. In Mar 2018, the cloud arm officially announced that it would strategically invest edge computing, and combine it with the advantages of cloud computing, big data, and AI. Developing edge computing could complement with Alibaba Cloud's benefits in security, storage, cloud computing, and AI.

Currently, the first level of Alibaba Cloud's edge computing is to transfer the edge IDC into the cloud infrastructure, which could cover the area between the cloud to places that are 100 kilometers far from the customers. The second level is the Mobile Edge Computing(MEC), which would be combined with the 5G network. As for the third level, it focuses on the customer and would be embedded into the Internet of the community, government institutions, enterprises, and families. The edge computing of Alibaba Cloud is moving forward level by level and aims to be applied to multiple scenarios.

With the increasing push toward Smart City initiatives, countries across the world are implementing edge computing solutions to enhance building security, home automation, parking, traffic management, and city asset management. Edge computing empowers IoT deployments in Smart Cities with low-latency and high bandwidth connectivity.

Smart Cities drives edge computing market to grow from USD 2.8 billion in 2019 to USD 9.0 billion by 2024, at a CAGR of 26.5% during the forecast period, according to a report published by MarketsandMarkets.

Numerous companies are adopting edge computing to enhance their Information Technology (IT) infrastructure and leverage the benefits of new technologies like 5G and IoT.

Alibaba Cloud is making a deeper push into smart cities in an effort to demonstrate the transformative impact of its technologies on society.

The city of Hangzhou, where Alibaba is based, is widely deploying the tech giant's Smart City project "City Brain 2.0," a cloud-powered and AI-driven urban traffic-management system that aims to boost the city's efficiency.

The core competency of the tech powerhouse is not only that it has a massive cloud infrastructure, but also it is the trend maker in critical areas such as urban-level artificial intelligence and enterprise digital transformation.

As the head of Alibaba Cloud's Edge Computing team, Yu revealed that Alibaba Cloud has completed a comprehensive coverage of 300+ edge computing nodes in more than 30 provinces in China. Furthermore, the cloud arm is deploying edge chips/devices, edge computing platforms (operating systems), urban edge middleware, urban edge applications, and services.

The AI inference chip Hanguang 800
The AI inference chip Hanguang 800

Back in September 2018, to support its cloud and IoT business, and the needs of different industries, the e-commerce giant said it would form a chip company to develop customized AI chips and embedded processors. Finally, Hanguang 800, the AI inference chip developed by Alibaba, was rolled out at Apsara Conference 2019. 

The self-developed chip is being implemented across many application scenarios within Alibaba Cloud, ranging from video classification to Smart City applications. The cloud arm says it previously used 40 traditional GPUs to process videos of central Hangzhou with a latency of 300ms. Now the mission requires only four Hanguang 800 with a lower latency of 150ms. 

According to the company, the AI inference chip can be used in autonomous vehicles, as well as smart city and logistics applications.

September 25, at the Apsara Conference, Zhang Jianfeng, president of Alibaba Cloud Intelligence, said that Alibaba Cloud gradually became an intelligent cloud provider from a cloud infrastructure provider.