Google Cloud - by Jiageng Mao
Google Cloud (https://cloud.google.com/) is a cloud computing platform which provides a suite of cloud computing services that runs on the same infrastructure that Google uses internally for its end-user products, such as Google Search and YouTube. Alongside a set of management tools, it provides a series of modular cloud services including computing, data storage, data analytics, and machine learning. We went to visit Google Cloud on the second day of our trip and listened to Max Ng’s sharing from his working experience.
Our speaker, Max Ng, is a software engineer in the data center of Google and works closely with engineers in Google Cloud. He graduated from CUHK and joined a startup after that. Then he came to the US and worked for Apple. Now he is a software engineer in Google and works on maintaining the infrastructure of data center.
At the beginning, our speaker showed us around the Google Cloud building. On the ground floor, there is a wall made of a large PCB board and chips with Google Cloud logo on it. At the resting place, there is a desk with a game machine embedding on which people can play Atari games. Then we went upstairs and saw many conference rooms in which engineers were working, as well as functional rooms like the mother’s room or cafeteria. We picked up some food and water and our speaker began his sharing.
Our speaker first shared with us his own working experience in HK and the US. After graduation, he came up with an idea what people can read recommended news from different channels in a single application so he co-founded a startup company. During this period he said that he put too much effort on the backend and didn’t pay enough attention to UI design. As a result, his app didn’t attract many people’s attention. After this failure, his friend recommended him to join Apple then he came to the US. He worked in Apple for several years and transferred his job in Google.
After that our speaker shared with us some knowledge on hardware and software design in the infrastructure of Google. He talked about the storage and communication system. A complete storage system includes cache-memory-hard drive/network storage-backup tapes and the most difficult problem is decreasing latency. He also talked about the difficulties in building a cloud computing platform, for example, how to get different types of programs to run stably on a single device. His sharing let us learn more knowledge about infrastructure design of a cloud computing platform.
At the Q&A section, our speaker shared with us his view on why Google cannot win over Amazon in the cloud service market. He said that Amazon started earlier and provided cloud service for most big companies. Cloud service made revenues mainly from big enterprises and once deciding to run their business on one platform, those enterprises are not likely to transfer to another new cloud platform. He also talked about his own opinion on whether Google Cloud will be into the Chinese market. He said the customer base and infrastructure is of great significance. If Google wants to set up cloud service in other regions, they will deploy servers in this region and compete with local companies on customers. And regulation of local government is also an important issue.