As the world’s undisputed leader in Internet search, Google has undeniable expertise in running data centers. After Amazon launched its cloud computing service in 2006, Google put that data center expertise to work by launching a cloud service of its own. Soon it became known as one of the “big three” public cloud vendors, along with AWS and Microsoft.
Depending on which analyst report you read, today Google is either the third or the fourth largest public cloud vendor, behind Amazon, Microsoft and (possibly) IBM. While its larger competitors tout their broad portfolios of services, Google has a narrower focus and specializes in meeting the needs of developers. Through its Web search and advertising businesses, the company has become particularly adept at handling big data, and it has transferred those capabilities to its cloud computing business as well, developing a solid reputation for prowess in analytics, artificial intelligence and machine learning.
History of Google Cloud Platform
Unlike Amazon and Microsoft, which both began their foray into the cloud with infrastructure as a service (IaaS) offerings, Google’s first public cloud service was its platform as a service (PaaS), called App Engine. App Engine first launched as a private preview for developers in April 2008. Initially, Google made the service available to just 10,000 users, but by the end of May, it had expanded to 75,000 developers with more than 80,000 more on the wait list. At that point, Google opened the service up to everyone. In the beginning, App Engine was free — and resources were limited — but Google began allowing developers to purchase resources beyond the free tier in 2009. However, Google left the “preview” label on App Engine until November 2011.
While developers were eager to try out Google’s cloud, some criticized App Engine for its lack of support for some of the most popular programming languages, most notably Java. Google remedied that in April 2009, when it officially added support for Java.
Google launched its second major cloud service — Cloud Storage — in May 2010, making its entry into the IaaS market. At that time, Google also expanded its support for enterprise users with Google App Engine for Business. Compute Cloud, Google’s competitor for AWS Elastic Compute Cloud and Microsoft Azure Virtual Machines, went live as a preview in June 2012.
Since then, Google has continued to expand its services and lower its prices. It prides itself on being the least expensive public cloud provider, and it also differentiates itself with its assortment of big data, machine learning and container management tools.
Google Cloud Platform Services
Google Cloud Platform’s portfolio of cloud services isn’t as extensive as Amazon’s or Microsoft Azure’s, but it offers some specialized tools for developers that are hard to find elsewhere. It organizes its cloud services into the following nine categories:
- Compute — includes Google’s original cloud service, the App Engine PaaS, as well as Compute Engine, Container Engine, Container Registry and the Cloud Functions serverless computing offering
- Storage and Database — includes Cloud Storage object storage, Cloud SQL with MySQL or PostgreSQL database options, Cloud Bigtable NoSQL database, Cloud Spanner relational database, Cloud Datastore NoSQL database and Persistent Disk block storage
- Networking — includes Virtual Private Cloud (VPC), Cloud Load Balancing, Cloud CDN content delivery, Cloud Interconnect for enterprise-grade connections to Google’s Cloud and Cloud DNS for Web domain serving
- Big Data — includes the BigQuery large-scale data warehouse, Cloud Dataflow for batch and stream data processing, Cloud DataProc with Spark and Hadoop capabilities, Cloud Datalab for analytics and visualizations, Cloud Dataprep for data preparation, Cloud Pub/Sub messaging, Genomics for genetic scientists and Google Data Studio for business reporting
- Internet of Things — includes Google’s Cloud IoT Core for secure device connection and management
- Machine Learning — includes Google’s Cloud Machine Learning Engine, as well as its APIs for job search, natural language, speech, translation, vision and video intelligence
- Identity and Security — includes Cloud IAM identity and access management, Cloud Identity-Aware Proxy, Cloud Data Loss Prevention API, Security Key Enforcement, Cloud Key Management Service, Cloud Resource Manager and the Cloud Security Scanner
- Management Tools — includes Stackdriver Overview, Monitoring and Logging, all of which support both Google Cloud Platform and AWS. This category also includes app optimization and deployment tools like Trace, Debugger, Cloud Deployment Manager, Cloud Endpoints, Cloud Console, Cloud Shell, Cloud Mobile App, Cloud Billing API and other Cloud APIs
- Developer Tools — Includes the Cloud SDK with its command line interface for Google Cloud Platform services, container tools, Cloud Source Repositories, Cloud Tools for Android Studio, Cloud Tools for IntelliJ, Cloud Tools for Powershell, Cloud Tools for Visual Studio, Cloud Tools for Eclipse, Gradle App Engine Plugin, Maven App Engine Plugin and the Cloud Test Lab for On-Demand Testing
Google boasts that its cloud services have customer-friendly pricing, and it claims that its prices are “on average 60% less for many compute workloads compared to other cloud providers.” The chart below, while far from exhaustive, offers an overview of the pricing and features for some of Google’s most popular cloud services. Please note, however, that pricing can and does change frequently.
Features and Costs of Popular Google Cloud Services
|App Engine||• PaaS with support for most language runtimes, frameworkds and third-party libraries.
• Free tier
• Highly scalable
• Open and flexible
• Powerful monitoring and management tools
|B1 instance classes at the Iowa data center starts at $0.05 per hour. Additonal fees may apply for services like data cells and search.|
|Compute Engine||• Highly scalable
• Custom machine types available
• Industry leadIng pricing
• Fast boots
• Minute-level incremental billing with automatic dlscounting
• Envlronmentally friendly data centers
|N1-standard-1 machinetypes at the Iowa data center start at $0.0475 per hour or $24.2725 per month.Preemptlble instances of thesame machine cost $0.01per hour or $7.30 permonth. Sustained use and committed use
discounts are available, as are custom machine types.
|Container Engine||• Built on Docker and Kubernetes
• Fast deployment
• Fully managed
• Integrateswith identity and access management services
|Upto 5 nodes are free; a cluster of more than 6 nodescosts $0.15 per hour or $109.50 per month,in addition to Compute Engine fees.|
|Cloud Storage||• Best-In-class performance and reliability.
• Low prices
• Seamless data lifecycle for active, nearline and colddata
• Built-in security
|Coldline Storage starts at just $0.007 perGB per month, while multi-regional storage costs $0.026 per GB per month.Data movement charges may also apply.|
• Low cost
• Serverless architecture
• Designed for analytics
• Fast performance
|$5 per TB for queries and $0.05 per GB for streaming inserts,in addition to applicable storage prices.|
|Cloud Machine Learning Engine||• Built on Google’s TensoFlow framework
• Support for thousands of users and TBs of data
• Deep learning and predictive analytics capabilities
|In the U.S.,model training costs $0.49 per hour per training unit and predictions cost $0.10 per thousand predictions plus $0.40 perhour, in addition to storage costs.|
Google Cloud Platform Advantages
Google’s biggest advantage when it comes to cloud computing is its experience as the world’s leading search engine. By powering Internet search, Google learned how to design and manage Web-scale data centers, and that expertise allows it to achieve high reliability and fast performance, as well as economies of scale that enable very low prices.
Those low prices are a key differentiator for Google. It has committed itself to price leadership in the public cloud, and offers a variety of discount programs and rightsizing recommendations to help customers keep costs low.
In addition, Google employs a vast number of developers itself, and that gives the company insight into what developers need and want from a cloud computing provider. Many of its cloud offerings are tailored for the needs of developers, particularly developers who are building cloud-native apps.
The company also touts its commitment to innovation and openness. It is a leader in the fast-growing area of data analytics and machine learning. It also boasts fast performance and easy set-up and deployment.
The Google Cloud Platform offers a diverse array of tools, and is considered particularly strong in data analytics.
When to Use Google Cloud Platform
Developers building cloud-native Web or mobile applications are often drawn to Google Cloud Platform, particularly the App Engine PaaS. The cloud computing service’s commitment to performance, innovation, openness and low prices make it popular with startups looking to scale rapidly. And while it is difficult to estimate and compare public cloud prices, Google’s commitment to price leadership make it a good choice for budget-conscious organizations of all sizes, and certainly for hybrid IT users.
Google also offers strong capabilities in big data analytics, machine learning and artificial intelligence. It had created and/or contributed to many of the most popular open source projects in this area, and it infuses its cloud offerings with that expertise. Those big data capabilities are augmented by Google’s inexpensive cloud storage, which makes this cloud platform cost effective for storing large volumes of data.
Similarly, Google is also working to distinguish itself in the area of container technology. It created the Kubernetes open source container management solution that is integrated into its cloud-based container tools, and it also has strong support for Docker.
In short, Google Cloud Platform does a few things particularly well. Gartner advises, “Evaluate GCP as if it were a specialized cloud platform for projects that play to these strengths. Many GCP customers have other primary cloud IaaS providers, but choose GCP for projects that are especially well-suited to the platform’s capabilities.”
As enterprises increasingly move toward multi-cloud strategies, it seems likely that many organizations will choose to use Google Cloud Platform to handle a portion of their cloud computing needs, particularly for new Web and mobile applications or big data analytics projects. Before choosing a cloud provider for your business, read our comprehensive cloud computing guide.
When Not to Use Google Cloud Platform
Few enterprises use Google Cloud Platform for all of their cloud computing needs. It simply doesn’t have the breadth and depth of offerings that AWS and Microsoft Azure have — yet. However, it has been making a concerted effort to reach out to enterprise customers and has been adding more services to its portfolio.
Google probably wouldn’t be the first choice for organizations that are migrating legacy applications to the cloud. It also might not meet the security and compliance requirements of some enterprises, like health care, financial or government organizations. In survey conducted by Cowen & Co.respondents pointed to Google’s IT support, cost and features as strengths, but listed compliance and security as its biggest weaknesses.
Also, although it has been adding more data centers, Google Cloud Platform doesn’t have as many regions as AWS or Azure. As a result, it lacks the global reach that these competitors can claim.
Still, Google continues to gain market share among enterprises and will likely remain a strong competitor in the public cloud market.