But why is this model so compelling nowadays? To understand why, we need to look at some history. The first wave of the trend that brought us towards cloud computing was colocation, which IT shops have been doing for decades. Instead of building costly capital intensive data centers, they can rent space in shared facilities. That frees up capital for more flexible uses than real estate. In the first decade of the 2000s, IT departments' need for efficiency drove them to use virtualization. The components of a virtualized data center match the parts of a physical data center; servers, disks and so on. But now there are virtual devices separately manageable from the underlying hardware. Virtualization lets us all use resources more efficiently and just like colocation, it lets us be more flexible too. With virtualization you still buy, house and maintain the infrastructure. So, you're still in the business of guessing how much hardware you'll need and when, setting it up and keeping it running. About 10 years ago, Google realized that its business couldn't move fast enough within the confines of the virtualization model. So, Google switched to a container based architecture, an automated elastic Third Wave cloud built from automated services. We'll explain exactly what containers are later in this course. In Google's internal cloud, services automatically provision and configure the infrastructure that is used to run familiar Google applications. Google has spent billions of dollars building this platform and making it resilient and efficient. Today, Google Cloud platform makes it available to Google customers.