Why Google Cloud Platform(GCP)?

Hua Shi
3 min readSep 22, 2020

--

Recently I am learning GCP and preparing the Associate Cloud Engineer Exam. I strongly believe that Cloud platforms will have a rapid upward trend in the future. There are several famous Cloud Platforms such as AWS, Microsoft Azure, IBM Cloud, and Google Cloud. Under the competition, those cloud platforms are getting mature and more functional. In the United States, the number of small businesses using cloud computing is expected to increase from the current 37% to 80% over the next six years.

So what is Cloud Computing?

Cloud computing can be defined as the computational power which completely resides on the cloud at all times. It is one of the latest innovations in the internet saga domain and primarily uses the Internet i.e. the Cloud, as its chosen method of delivery.

What are the advantages of GCP?

  • Lower price — Compare other Could Platforms, GCP’s prices are much affordable.
  • Speed— Faster Cable system can provide the speed which is up to 10Tbs.
  • Big Data & Data Analytics — There are some excellent cloud applications are very good to deal with big, massive, and real time data.

Before I always analyzed different datasets locally with Python. I spent 3 days running my chatbot model locally, and it broke due to a lack of storage, then I felt so exhausted. That is why I decided to learn GCP! Here is a short video of GCP Machine Learning Introduction. You will love it for sure!

When I was struggling with my Chatbot project, I watched this video and I told myself — yes this is exactly what I want.

GCP is really easy to use and at low prices! Yes, GCP is not free, but you can \ gain $300 credit as a gift from GCP after registering an account and you can use it within one year. For a project, the $300 is good enough to finish your task (for a single project use).

Our (team) chatbot project is based on the sequence to sequence model (RNN). This project requires a big scale data and TensorFlow environment. Since the data is text data, Natural Language Processing (NLP) is a ‘must’ part. GCP provides the auto-NLP application for the users. What we need to do is just to load the data and wait for the result. So here we don’t need to manually process the text data with NLTK.

Besides, we know that for RNN, the larger data we have, the better result we will get. If I train the model locally, my laptop cannot offer better storage, GPU, CPU, and TPU. The other “headache” problem is — “Time Cost”. It will run for almost more than 3 days! During those days I could not do any other projects, because the “heart” of my laptop (CPU) is occupied by the chatbot project. GCP offers a better environment to bring a “solid” power! And I do not need to worry about the storage of my laptop. The only thing I need to do is to wait for my result!

Lots of companies love cloud computing because not only it can help companies to reduce hardware costs but also employees can work anywhere with any computers. Furthermore, GCP stands out its data analytics. For example, BigQuery can process massive data at high speeds.

One of GCP’s products is Google Kubernetes Engine (GKE) which is widely used by many companies such as Pizza Hut, Alpha Vertex, Sky Italia and Tokopendia, etc. GKE can speed up app development with strong security and easily handle auto-repair and auto-upgrade.

If you are interested in GCP, you can visit their official website and there are a lot of useful documents. Also, Udemy and Coursera offer relevant online courses.

--

--

Hua Shi
Hua Shi

Written by Hua Shi

Data Engineer /Data Analyst /Machine Learning / Data Engineer/ MS in Economics

No responses yet