The Google On Board 2018 meeting promised to be massive, matching up to the company’s level. The location: the historical River Plate stadium.
Before discussing the content, we have to say that there were a few things that were not up to the event’s importance, such as the lacking preparation against the low temperatures, so much so that almost half of the attendees left after the break. Initially, hot beverages were limited to two per person, but that restriction was then dropped off due to the harshness of the weather. Many participants showed their discontent in social media, and people in other countries in Latin America echoed the Argentina experience.
(#gHelp it’s super cold here at the #GoogleOnboard in #Argentina, did you know that is winter in this country? I mean, for a 6 hours event at an outdoor stadium!)
(Oh man, that #GoogleCloudOnBoard felt ugly in Mexico)
Coming back to our subject of interest, let us discuss the technical side of the event.
Google on Board 2018: Technical Content
Google claims to have the fastest and most powerful backbone available connecting each one of the services in their network, and a processing speed undoubtedly superior to that of their competitors. With YouTube, Google Chrome, Gmail and other services being based on the Google Cloud, we can appreciate how robust and dependable their infrastructure is, since it is always available to users.
Google Cloud (like all Google products) aims to be easy to use, understand and organize. It enables organization by “projects,” so that we can have all of our resources grouped in one place. Besides the fact that each resource is only three clicks away (a great advantage for those who are cloud computing beginners), there is also the benefit of having the approximate cost of each resource always at display in the administration panel.
Four services stand out in the infrastructure offering:
1 – Compute Engine: Virtual machines that, unlike those of other providers, can be customized for RAM space, processor, and video card.
2 – App Engine: A PaaS framework thanks to which applications can be quickly deployed with minimal effort, enabling to focus on coding instead.
3 – Container Engine: Kubernetes, the containers orchestrator created by Google -there isn’t much more to say. It helps to create a dynamic, elastic, highly available flaw-tolerant infrastructure by easily using Docker.
4 – Cloud Functions: Google’s serverless service, which leverages the power of infrastructure to enable a transparent escalation. It is still in development stage for new features.
Surely, one thing Google is proud of is their Machine Learning services. They have created one of the most frequently used frameworks for this purpose: TensorFlow. They took one step further and developed the Tensor Processing Unit, a chip specifically designed for Machine Learning that is being extensively used in various services. They are ready to implement it in different levels.
Google Cloud provides solutions with different levels of complexity, including services that even allow infrastructure experts to create and quickly train their algorithms. It also provides trained Machine Learning services, and the possibility to translate text into different languages and to analyze images, text and audio. These resources are applied in their own products, too: Google Translate, Google Photos, virtual assistants, chatbots.
In sum, Google wants to secure their position in Latin America through massive events which, despite being poorly organized, encourage developers to use their platform and also to specialize or become certified in data engineering. In our view, this is a challenging effort, even for a company like Google. They need to improve certain aspects before they begin to be considered a reliable choice in the future.
PS: Google, we want you to know that the food court was NOWHERE to be found =(.
(It is not allowed to enter with food or drinks, but don’t worry: we offer food and drinks all day long, and you’ll also have a food court at your disposition. We are waiting for you! Sincerely, the Google Cloud Platform team)